Citation
Cross reactive antigens and lectin as determinants of host specificity in the Rhizobium-clover symbiosis

Material Information

Title:
Cross reactive antigens and lectin as determinants of host specificity in the Rhizobium-clover symbiosis
Creator:
Dazzo, Frank Bryan, 1948- ( Dissertant )
Hubbell, David H. ( Thesis advisor )
Smith, Paul H. ( Reviewer )
Bleiweis, Arnold S. ( Reviewer )
Hoffmann, Edward M. ( Reviewer )
Preston, James F. ( Reviewer )
Charudattan, Raghavan ( Reviewer )
Fry, Jack L. ( Degree grantor )
Place of Publication:
Gainesville, Fla.
Publisher:
University of Florida
Publication Date:
Copyright Date:
1975
Language:
English
Physical Description:
x, 68 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Agglutination ( jstor )
Antibodies ( jstor )
Antigens ( jstor )
Antiserum ( jstor )
Clover ( jstor )
Cross reaction ( jstor )
Infections ( jstor )
Lectins ( jstor )
Rhizobium ( jstor )
Root hairs ( jstor )
Clover ( lcsh )
Dissertations, Academic -- Microbiology -- UF
Microbiology thesis Ph. D
Rhizobium ( lcsh )
Symbiosis ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Abstract:
The basis for host specificity in the Rhizobium-Trifolium (clover) symbiosis was investigated. Cross reactive antigens of clover roots and R. trifolii were detected on their cell surfaces by tube agglutination, immunofluorescent and radioimmunoassay techniques. Anticlover root antiserum had a higher agglutinating titer with infective strains of R. trifolii than with noninfective strains. The root antiserum previously adsorbed with noninfective R. trifolii cells remained reactive only with infective cells, including infective revertants. When adsorbed with infective cells, the root antiserum did not react with either infective or noninfective cells. Other Rfuzob-cum species incapable of infecting clover did not demonstrate surface antigens cross reactive with clover. Radioimmunoassay indicated twice as much antigenic cross reactivity of clover roots and R. trifolii 403 (infective) as compared with R. trifolii Bart A (noninfective) . Immunofluorescence- using anti-R. trifolii 403 (infective) antiserum was detected on the exposed surface of the root epidermal cells and diminished at the root meristem. The immuno fluorescent cross reaction on clover roots was totally removed by adsorption of anti-R. trifolii (infective) antiserum with encapsulated infective cells but not with noninfective cells. The cross reactive capsular antigens from R. trifolii. strains were extracted and purified. The ability of these antigens to induce clover root hair deformation was much greater when they were obtained from infective as compared with noninfective strains. The cross reactive capsular antigen of R. trifolii 403 was characterized as a high molecular weight (>4.6 x 10 daltons) , amorphous, B-linked, acidic heteropolysaccharide containing 2-deoxyglucose, galactose, glucose, and glucuronic acid. A soluble, nondialyzable substance (clover lectin) capable of binding to the cross reactive antigen and agglutinating only infective cells of R. trifolii was extracted from white clover seeds. This lectin was sensitive to heat, pronase, and trypsin. Inhibition studies indicated that 2-deoxyglucose was the most probable haptenic determinant of the cross reactive capsular antigen capable of binding to the root antiserum and the clover lectin. Infection and nodulation of white clover roots by R. trifolii was inhibited by 2-deoxyglucose at concentrations of at least 30 mM. The adsorption of infective and noninfective R. trifolii cells to white clover root hairs was examined. After 12 h of incubation, four to five times more infective cells than noninfective cells were firmly adsorbed to root hairs. A model is proposed to explain host specificity based on the preferential adsorption of infective versus noninfective cells of R. trifolii.1 on the surface of clover roots by a cross bridging of their common surface antigens with a multivalent clover lectin.
Thesis:
Thesis (Ph. D.)--University of Florida, 1975.
Bibliography:
Bibliography: leaves 63-67.
Additional Physical Form:
Also available on World Wide Web
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Frank Bryan Dazzo.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
028525093 ( AlephBibNum )
08819198 ( OCLC )
ABU4299 ( NOTIS )

Downloads

This item has the following downloads:


Full Text









CROSS REACTIVE ANTIGENS AND LECTIN AS
DETERMINANTS OF HOST SPECIFICITY IN.
THE RHIZOSIUM-CLOVER SYMBIOSIS












By

FRANK BRYAN DAZZO


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY












UNIVERSITY OF FLORIDA

1975


















To my wife, Olga











Exploration

We shall never cease from exploration,
And the end of all our exploring
Will be to arrive where we started
And know the place for the first time.

Anonymous








ACKNOWLEDGEMENTS


The author would like to express his deep apprecia-

tion to the chairman of his committee, Dr. David H. Hubbell

for his constant guidance, concern, interest, and support.

He would like to thank his committee members Drs. Raghavan

Charudattan, Arnold S. Bleiweis, Edward M. Hoffmann, and

James F. Preston for their assistance and suggestions and

especially Drs. Paul H. Smith and Charles F. Eno for many

years of friendship and support.

He would like to thank Mr. Manuel Mesa, Mr. James Struble,

and Dr. Willis Wheeler for their professional assistance and

especially Ms. Carolyn Napoli, who has been a pleasure to

work with throughout this study. He would also like to

thank Ms. Napoli for providing the electron photomicrograph

of Rhizobium tiLdolii presented in Fig. 19. The author thanks

the many scientists listed in Table 1 who provided the

Rhizobium cultures used in this study.

The author wishes particularly to express his loving

gratitude to his wife, Olga, for her encouragement and

understanding which made this dissertation possible.

This research was supported by the National Science

Foundation Grant No. GB-31307 and a Grant-in-Aid for

Research from Sigma Xi, the Scientific Research Society of

North America.














TABLE OF CONTENTS

Page


ACKNOWLEDGMENTS ........................................ iii

LIST OF TABLES ......................................... v

LIST OF FIGURES ....................................... vi

ABSTRACT ...............................................viii

INTRODUCTION ..... ............. .... ......................... 1

LITERATURE REVIEW ..................................... 2

MATERIALS AND METHODS .................................. 5

RESULTS AND DISCUSSION ................................. 18

LITERATURE CITED ....................................... 63

BIOGRAPHICAL SKETCH .................................... 68














LIST OF TABLES


Table Page

1. Sources and Infectivity Character-
istics of Rhizobium Strains .................... 6

2. Agglutination of Rhizobium tniotii with
Rabbit Anti- TAiL6oeim )epens Root Antiserum ... 19

3. Reactivity of Anti-TAi6otLim tepens Root
Antiserum and Rhizobi m ttLioZii Cells
using Indirect Immunofluorescence .............. 20

4. Deformation of Tai6goium fAaqgifeiam Root Hairs
Induced by Capsular Material from
Rhizobium triAotii Strains ..................... 42

5. Agglutination of Rhizobium Cells by
TA.iLotium Azpens Seed Extract .................. 46

6. Binding of Clover Lectin to Rhizobium Cells ...... 50

7. Effect of Various Treatments on Agglutina-
tion of Rhizobium .tiLotLii 403 by Clover
Seed Extract .................................. 51

8. Inhibition of Clover Lectin-Mediated Cell
Agglutination by Various Carbohydrates ......... 53

9. Inhibition of Infection and Nodulation of
Thifoaium repeAs with Rhizobi-m t-i6oti.i
403 by 2-Deoxygluccse .......................... 55

-0. Adsorption of Rhizobium ti'oZtii Cells to
Ti6otlium kepens Root Hairs .................... 56













LIST OF FIGURES


Figure Page

1. Immunofluorescence of Rhizobium ttifolii
403 (infective) using anti-Ti.AotLim repen6
root antiserum ................................. 22

2. Radioimmunoassay of anti-TAijoZium erpens.
root antiserum bound to Rhizobium ttiLotZi
403 (infective) and Bart A (noninfective)
cells .......................................... 25

3. Immunofluorescence of a T/riolium tepenb root
using anti-Rhizobium t i otii 2S-2 antiserum ... 28

4. Rhizobium tifoZlii 403 capsular antigen
dissolved in water (2 mg/ml) ................... 31

5. Reaction of Rhizobium tLA(oZii 403 capsular
antigen with DEAE-dextran ...................... 31

6. Reaction of Rhizobium thtdolii 403 capsular
antigen with anti-Tjlctiutm repen root
antiserum ...................................... 31

7. Reaction of Rhizobium t'Zioiti 403 capsular
antigen with TtiLotium Aepens seed extract ..... 31

8. Quantitative precipitin curve of Rhizobium
ttiotii 403 capsular antigen with anti-
Ttifoiium Repens root antiserum ................ 32

9. High pressure liquid chromatography of the
Rhizobium tLiSolii 403 capsular antigen on
Bio-Glas 2500 .................................. 34

10. Gas-liquid chromatography of trimethylsilylated
sugars in an acid hydrolysate of Rhizobium
tLi6olii 403 capsular antigen .................. 36

11. Infrared spectra of Rhizobium tAiiotil 403
and Bart A capsular antigens ................... 38

12. The effect of pH on the immunofluorescent cross
reactivity of encapsulated cells of Rhizobium
tti6ofii 403 with anti-TAi6otium tepens
root antiserum ................................. 40


vi








LIST OF FIGURES--Continued


Figure Page

13. TrifoLiium fragifeg um root hairs ................. .45

14. TAidotium r6Lagig6eum root hair deformation
in the presence of the capsular material
from Rhizobium Ltitotii 403 (100 ug/ml) ....... 45

15. TaLijoium fAagi6eaum root hair deformation
in the presence of the capsular material
from Rhizobium tfijoii Bart A (100 ug/ml) .... 45

16. Rhizobium tLi6otii 2S-2 cells suspended in
saline ....................................... 48

17. Agglutination of Rhizobium tt6iolii 2S-2 cells
by the TAiootium Aepens seed extract .......... 48

18. Adsorption of Rhizobium tfioZii 403 cells on
a TLid otium Repens root hair .................. 58

19. Electron photomicrograph of an ultrathin sec-
tion of Rhizobium Lijotii NA-30 in association
with a Ti6otLium fragifeaum root hair ......... 58

20. Schematic diagram of the proposed cross-bridging
of the cross reactive antigens of Rhizobium
triLotii and TZifotium root hair with
a clover lectin .............................. 60













Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



CROSS REACTIVE ANTIGENS AND LECTIN AS
DETERMINANTS OF HOST SPECIFICITY IN
THE RHIZOBIUM-CLOVER SYMBIOSIS


By

Frank Bryan Dazzo

December, 1975


Chairman: Dr. David H. Hubbell
Major Department: Microbiology

The basis for host specificity in the Rhizobium-

TAri6olium (clover) symbiosis was investigated. Cross

reactive antigens of clover roots and R. tlifoZii were

detected on their cell surfaces by tube agglutination,

immunofluorescent and radioimmunoassay techniques. Anti-

clover root antiserum had a higher agglutinating titer with

infective strains of R. tAi- oZii than with noninfective

strains. The root antiserum previously adsorbed with non-

infective R. tifoLZii cells remained reactive only with

infective cells, including infective revertants. When

adsorbed with infective cells, the root antiserum did not

react with either infective or noninfective cells. Other

Rhizobium species incapable of infecting clover did not

demonstrate surface antigens cross reactive with clover.


viii







Radioimmunoassay indicated twice as much antigenic cross

reactivity of clover roots and R. ttiLotii 403 (infective)

as compared with R. tAifoZii Bart A (noninfective). Im-

munofluorescence-using anti-R. tAidotii (infective) anti-

serum was detected on the exposed surface of the root epi-

dermal cells and diminished at the root meristem. The

immunofluorescent cross reaction on clover roots was totally

removed by adsorption of anti-R. tnifotii (infective) anti-

serum with encapsulated infective cells but not with non-

infective cells.

The cross reactive capsular antigens from R. thiLAoii

strains were extracted and purified. The ability of these

antigens to induce clover root hair deformation was much

greater when they were obtained from infective as compared

with noninfective strains. The cross reactive capsular

antigen of R. tAifotii 403 was characterized as a high

molecular weight (>4.6 x 106 daltons), amorphous, --linked,

acidic heteropolysaccharide containing 2-deoxyglucose, gal-

actose, glucose, and glucuronic acid.

A soluble, nondialyzable substance (clover lectin)

capable of binding to the cross reactive antigen and agglu-

tinating only infective cells of R. ttLiotii was extracted

from white clover seeds. This lectin was -sensitive to heat,

pronase, and trypsin. Inhibition studies indicated that

2-deoxyglucose was the most probable haptenic determinant

of the cross reactive capsular antigen capable of binding

to the root antiserum and the clover lectin. Infection







and nodulation of white clover roots by R. tAihotii was

inhibited by 2-deoxyglucose at concentrations of at least

30 mM. The adsorption of infective and noninfective R.

tAifolii cells to white clover root hairs was examined.

After 12 h of incubation, four to five times more infective

cells than noninfective cells were firmly adsorbed to root

hairs. A model is proposed to explain host specificity

based on the preferential adsorption of infective versus

noninfective cells of R. tLidolii on the surface of clover

roots by a cross bridging of their common surface antigens

with a multivalent clover lectin.














INTRODUCTION


The N2 fixing Rhizobium-legume symbiosis is character-

ized by a high degree of host specificity controlled by the

bacteria and the plant. Expression of host specificity is

an early event which occurs prior to the formation of

infection threads within the root hairs of the host (29).

The purpose of the study reported here was to determine

the basis for host specificity in the Rhizobium triLolii-

TaifoZium (clover) symbiosis. It was hypothesized that

R. ttAiotii cells have specific surface antigens which are

involved in the adhesion of the bacteria to the cross re-

active surface antigens of the clover root hair cell wall

with the aid of a clover lectin. This association was

chosen for study because of the small-seeded nature of the

macrosymbiont which was ideal for microscopic studies of

infection and the availability of mutant strains of R.

zifdolii which have lost the ability to infect clover.


- 1 -













LITERATURE REVIEW

Cross Reactive Antigens

The importance of common or cross reactive antigens

in the microbial invasion of a host is receiving increasing

recognition. Antigenic similarities between vertebrate

hosts and microbes have been implicated in the pathogenesis

of rheumatic fever (55), human ulcerative colitis (41), rat

glomerulonephritis (34), and rat arthritis (6). In these

cases, pathogenesis may involve immunological tolerance of

the cross reactive antigens by the host or deposition of

cross reactive cytotoxic antibodies on the host tissue.

Common antigenic substances between invasive microbes

and their plant hosts have also been found. These anti-

genically similar cell constituents possibly underlie

host-pathogen compatibility based on their correlation with

disease development. According to one theory (16, 17), a

strong common antigen relationship between a plant host and

a pathogen might result in the least disruption of cellular

function between a pathogen and its host during infection

with consequent success in disease development. Common

antigen relationships have been implicated in the patho-

genesis of flax rust by Melampor.a. ini (18), angular leaf

spot in cotton by Xanthomonas malvaceaLum (17, 45), black

rot of sweet potatoes by Cexatocysti4 fimbtiata (17), common

smut of corn by Ustitago maydiz (52), crown gall tumor of

2 -




- 3 -


tobacco by Ag obacteAium tumefaciten (8), and wilt of cot-

ton by FuAacium and Vaetic.lltium (9). Host specificity is

sharply defined for all of these phytopathogens except A.

tumedaciend which has a wide host range. Cross reactive

antigens have been found between 8 legumes and 3 species

of Rhizobium (10). However, there was no correlation be-

tween the numbers of common antigens (immunoprecipitin bands

in Ouchterlony plates) and the ability of the bacteria to

infect their respective legume hosts.

Lectins

A lectin is a non-antibody protein or glycoprotein

capable of specific interaction with carbohydrates. Lectins

have been isolated from a variety of plants and non-vertebrate

animals. Many lectins are phytohaemagglutinins since they

can agglutinate erythrocytes by binding to their surface

carbohydrate components.

Interactions between legume lectins and Rhizobium have

been suggested (1, 48) as well as documented (4, 5, 7, 15, 22).

Hamblin and Kent (22) showed that phytohaemagglutinin

(Phaseolus lectin) could bind to R.phaseoa i and that this

lectin was present in the seeds, nodules, and on the roots

below the nodules of Phaseolus vulgatis. Bohlool and

Schmidt (4) demonstrated a high correlation between the

binding capacity of soybean lectin to Rhizoblum cells and

the ability of these bacteria to nodulate soybean. They

proposed that the legume lectin may serve as the basis for

host specificity by interacting specifically with a poly-

saccharide on the surface of the Rhizobium cell. The




4 -



lectin binding site on the Rhizobium cell and the means

whereby the lectin binds to the plant roots were not

examined. However, some strains of Rhizobium do not bind

to lectins obtained from the legume host that they nodulate

(4, 5), and other strains of Rhizobium incapable of nodu-

lating a certain legume still bind to the lectin from

that host (5, 15). Thus, interactions between legume lectins

and Rhizobium cells may not always account for the specificity

expressed by the nodule bacteria for their respective

legume hosts (15).













MATERIALS AND METHODS

Strains of Rhizobium

The sources, infectivity, and legume hosts of the

Rhizobium strains are listed in Table 1. Infectivity is

defined as the formation of root hair infection threads in

small-seeded legumes using glass slide assemblies (20), or

the production of root nodules on large-seeded legumes

planted in cellophane pouches (51). The spontaneous infec-

tive revertants of R. t'iAoLii (BA-L, BA-S, and 0435-21)

were isolated from nodules of TAi6otium Repend inoculated

with the corresponding noninfective strains.

Preparation of Antigens

Bacterial cells were grown on a modified Bergersen's

chemically defined medium, harvested, and sonicated as

previously described (14). Root antigens of T. repens var.

Louisiana Nolin and T. raagige&um var. Salina (hereafter

called white and strawberry clover, respectively) were

prepared. Seeds (50 g) were surface sterilized, spread on

water agar plates, overlayed with sterile stainless steel

wire mesh, inverted, and cold treated (38). The seeds

germinated through the wire mesh into humid air at 22 C.

After 3 days, seedling roots were excised along the wire

mesh with razor blades and frozen in liquid N2. Roots were

macerated by grinding and thawing in 30 ml phosphate buffered

saline (PBS, 0.05 M K2HPO4-KH2PO4, 0.15 M NaC1, 0.001 M MgSO4,

5 -





- 6 -


'-4 0
0
-4 N

) 0 0
Cfl C~l )
.. r-A N
~Ia = a f
04 *- Cd d






























+zz






ca r m .,
ulQ 0




















co ~ ~ U I~ I m a C .Z C

.9
E-1

0 Cd








UI I-


LOr
*i94.







c I H






Cd LL N N I
U, Cl d j~ I 0 o

4.3I I LA A LAo ( Z NE







0d Cf dc*


'H
0)r
'H~
.0i




- 7 -


C C $4
r- d V4
N 0 4 a)

N p .0 u m

a 3 3 z X M
H 0
a) U
S S o
s u h m


hr 1- Q' 5 3


5


E Y
ci 9-

E49
r

' E
ctf

) -3
C3a
49 .9
(S 0-


ci


a
T :
-(U

0
01
ci
ci

-a
ci
atl


ci





-t

ci


49:




ci
-(
ci


ci


di
ci
<
-C






ci
ci



SS


<
E
0
e
-s,


a)
a)


+ + + + + I + I + ++ + +


0
C 0

< i 0 -4
- -q


di
ci

ci
di-
to
ci C
'9Z '9
Cdi
o ci


.-, 49


49

o~


--4


l 0


oa 01 ( N H(c

N ON -N ) u
(N (N

49:
.9

49


-9
0
tt)

4 *
ci
-c 04
U, 0
a; w


>1


*H



U-I

-r


0







'4

0






'U
0




.,-
a)












cn











0
0o

4J





0+




0 ro
4
'U
a









u.n
.-.(










c
a)
$4





e-4






-o n


(0









o ^b
N Q




- 8 -


pH 7.2) containing 9.05 M sodium ascorbate (9) and 30 g of

insoluble polyvinylpyrrolidone (PVP, Bio-Rad Laboratories,

Richmond, Calif.) which had been previously washed with PBS

8 times to remove 280 nm absorbing material (13). The slurry

was stirred at 0 C for 20 min, then centrifuged at 27,000 x g

for 30 min. The supernatant was decanted, concentrated

6-fold by dialysis at 4 C against 20% (w/v) polyethylene

glycol (PEG) dissolved in PBS (14), dispensed in 2 ml

aliquots, and stored at -70 C for future use. Particulate

antigens were also prepared by mascerating seedling roots

in PBS containing 20% soluble PVP (40,000 av. mol. wt.,

pharmaceutical grade, Sigma, St. Louis, Mo.) and then

ultracentrifuging at 104,000 x g for 3 h. The pellet was

washed twice with PBS, ultracentrifuged between washes, and

then used as particulate antigens for immunization.

Preparation of Antiserum

Rhizobium and clover root antigens were emulsified with

equal volumes of Freund's incomplete adjuvant and used to

prepare antiserum in virgin New Zealand white rabbits ac-

cording to the immunization schedule previously described (14).

S.eep anti-rabbit y globulin antiserum and fluorescein

isothiocyanate (FITC) labeled y globulin fraction of goat

anti-rabbit y globulin were purchased from Nutritional

Biochemicals Co., Cleveland, Ohio, and Difco Laboratories,

Detroit, Michigan, respectively.

Serological Techniques

Tube Agglutination

Tube agglutination of Rhizobium cells was performed




- 9 -


using saline (0.15 M) as the diluent (50). Eight strains

of R. trhiolii were examined, including 4 infective strains

(25-2, 403, T37, and 0435) and 4 noninfective mutants

(2L, Bart A, Bio-9, and 0435-2).

Immunofluorescence

Both the direct and the indirect immunofluorescent

tests were run according to standard procedures (21).

Controls for detection of autofluorescence, nonspecific

staining, and normal serum reactivity were included. Details

of the optical system employed were described elsewhere (15).

Immunoprecipitation

A quantitative precipitin test using the purified

capsular antigen of R. trifolii strain 403 and rabbit

anti-white clover root antiserum was performed according

to Nowotny (40). The protein contents of the washed im-

munoprecipitates were determined with the Folin phenol

reagent (31). The supernatants were analyzed for excess

antigen or antibody using the capillary precipitin test.

Radioimmunoassay

For the radioimmunoassay, 10 pg (protein) of the

- globulin fraction (21) of sheep anti-rabbit y globulin

were iodinated with 2 mCi of Na I (Amersham Searle,

Arlington Heights, Ill.) using the chloramine T method

(25). The iodinated protein was separated from free iodine

by gel filtration on a 1 x 10 cm Sephadex G-50 column

(Pharmacia, Uppsala, Sweden). R. ti6otii cells (strains

403 and Bart A) were suspended in PBS to a final density

of 1.6 x 10 cells per ml as determined by direct counting




- 10 -


with a Petroff-Hausser chamber. Rabbit anti-white clover

root antiserum (0.01 ml) was added to 0.05 ml of the cell

suspensions run in duplicate, incubated at 37 C for 30 min,

0 C for 30 min, and then brought up to 1 ml volume with

PBS. The suspensions were washed twice with PBS by

centrifugation at 1,000 x g for 10 min each. This

procedure of adding cross reactive antibody, incubation and

washing was repeated twice. The washed cells adsorbed

with the cross reactive antibody were incubated with the
125I labeled goat anti-rabbit y globulin added in 0.06 ml

increments. The reaction mixtures were incubated for 30 min

at 37 C, followed by 30 min at 0 C. Following washing

of cells as before, the pellets were counted for y emissions

in a Model E116A-0 y counter (Nuclear Equipment Co.,

Farmingdale, N. Y.) This procedure of adding the 125I

labeled anti-y globulin was repeated until no further

increase in antibody binding occurred, indicating antibody

saturation. Tubes lacking anti-clover root antiserum were

processed to quantitate non-specific binding of antiglobulin

to cells, the inner wall of the tubes, and background

- radiation.

Adsorption of Antiserum

One ml of antiserum was incubated at 0 C with 0.05 ml

packed cell volume for 15 min with constant mixing, and then

centrifuged at 1,500 x g for 15 min. The supernatant was

readsorbed twice with fresh cells. The adsorbed antiserum

was unreactive with the adsorbing cells as determined by

indirect immunofluorescence. Encapsulated cells of R.




- 11 -


t~niolii strain 403 used for adsorption of cross reactive

antiserum were separated from unencapsulated cells by

differential centrifugation in PBS at 1,500 x g for 20

min. Clean, reliable separations were achieved under

these conditions. Unencapsulated cells sedimented faster

than encapsulated cells, forming a pellet under the

diffuse layer of encapsulated cells. This upper layer

was drawn off and centrifuged at 27,000 x g for 15 min

to pellet the encapsulated cells.

Analytical Studies of Rhizobium Capsular Antigens

Purification

The capsular antigens of the R. taiJoZii strains were

purified by procedures using cetylpyridinium chloride (40),

and hexadecyltrimethyl ammonium bromide (11) as completing

agents. The material was lyophilized and stored in a des-

iccator for future use.

Precipitation and Immunofluorescence

The capsular antigen of R. titoliiL 403 (hereafter called

403-AC) was tested for reactivity with 1% soluble

diethylaminoethyl-dextran (DEAE-dextran) and 1% dextran

sulfate (Pharmacia [191), clover lectin (discussed later),

and homologous and heterologous cross reactive antibody in

capillary tubes. Some 403-AC was spread on microscope

slides and tested for reactivity with rabbit anti-white

clover root antiserum and anti-R. thijotii 403 antiserum

using the indirect immunofluorescent technique.

Electrophoresis

The capsular materials (2 mg/ml) from R. tuioecii




- 12 -


403, and Bart A (hereafter called BAC), were spotted on

cellulose polyacetate strips (Sepraphore III, Gelman Co.,

Ann Arbor, Mich.) and electrophoresed for 1 h at 22 C

under constant voltage (300 V) at pH 8.6 (0.05 M sodium

barbital). The strips were dried, stained in 1% alcian

blue in 0.01 N HC1 for 10 min, and then washed in 0.01

N HC1 overnight.

Gel Permeation Chromatography

A sample of 403-AC was hydrated from the lyophilized

state and was passed through a 0.20 pm filter. An average

molecular weight was determined by gel permeation (3) using

a high pressure liquid chromatograph (Model 202/401 Waters

Associates, Milford, Mass.) A sample of 403-AC was

chromatographed in two stainless steel columns (3.2 x 600 mm)

connected in series and packed with porous glass beads

(AX through FX, Waters Assoc., and Bio-Glas 2500, Bio-Rad

Laboratories) deactivated with 1% PEG, and was detected by a

refractive index monitor. Degassed water was the solvent.

The columns were calibrated by determining the elution

volumes of soluble dextrans (5 x 10 to 2 x 10 daltons,

av. mol. wt., Sigma) and purified plant viruses (Brome-
6 7
grass mosaic, 4.6 x 10 and tobacco mosaic, 3 x 10 daltons,

courtesy of Dr. E. Hiebert, Dept. Plant Pathology, Univ. Fla.)

Biochemical Component Analysis

Samples of 403-AC and BAC were acid hydrolyzed (11),

converted to the volatile trimethylsilylated derivatives

with Tri-Sil (Pierce Chem. Co., Rockford, Ill. [43]), separa-

ted and detected by gas-liquid chromatography employing







flame ionization detection as previously described (38).

Sugar derivatives were identified by comparison with

retention times of authentic standards (Nutritional

Biochemicals Co.) The 4-0-methyl, D-glucuronic acid (24)

was a gift from Dr. F. Loewus, Dept. Biology, State Univ. of

New York (Buffalo, N. Y.); a-D glucuronic acid was a gift

from Calbiochem, La Jolla, Calif. The neutral sugar and

uronic acid content of 403-AC was quantitated colorimetrically

by the anthrone and carbazole methods, respectively (28).

Amino acid and amino sugar analyses of 403-AC hydrolysates

were conducted on a JOEL Model JLC-6AH automated amino

acid analyzer (JOEL, Inc., Cranford, N. J. [49]).

Diffraction and Spectral Analyses

X-ray diffraction and infrared spectral analyses of

403-AC and BAC were performed as previously described (38).

A UV absorption spectral analysis of 403-AC dissolved in

water (2 mg/ml) was performed using Suprasil cuvettes

(Precision Cells Co., Hicksville, N. Y.) and a Beckman DBG

grating spectrophotometer.

Antigenic Cross Reactivity of R. ttilotii 403 After
Enzymatic Digestion

Cells were fixed on microscope slides, then incubated

4 days at 22 C in the dark with the highest purity grade

of pronase, trypsin, lysozyme, deoxyribonuclease, ribonuclease,

pectinase, cellulase, and phospholipase dissolved in their

appropriate buffers (Worthington Biochemical Corporation,

Freehold, N. J. [53]). Cells were also treated with 7 M

urea, 1% Triton X-100 (Sigma), 1% Tween-80 (Difco), 0.5 M


- _L 1 -




- 14 -


sodium periodate, aqueous HC1 (pH 3, 4, 5, and 6), deionized

water (pH 7), and aqueous NaOH (pH 8, 9, 10, 11, and 12).

Cells plus the enzyme buffers were also examined. After

incubation, cells were washed with deionized water and tested

for antigenic cross reactivity using anti-white clover root

antiserum by the indirect immunofluorescent technique.

Biological Assay

The effect of the capsular material (100 ig/ml, [23]) on

strawberry clover root hair development was determined as

previously described (46). Root hairs along the 4 optical

median planes of two seedling roots per capsular preparation

were examined by phase contrast microscopy. Hairs were

counted, averaged, placed into straight, branched, moderately

curled (<3600), or markedly curled (>3600) categories (54),

and the data evaluated statistically.

Clover Lectin Studies

Extraction

White clover seeds (50 g) were ground to a fine powder

and mixed with 150 ml PBS plus sodium ascorbate (0.05 M) with

further grinding. Washed, insoluble PVP (20 g) was added,

the slurry was stirred at 4 C for 1 h, filtered through PBS-

washed cheese cloth, and centrifuged at 27,000 x g for 1 h.

The upper lipid layer was removed, and the remaining

supernatant was passed through 0.20 pm membrane filters

(TCM-200, Gelman). This seed extract (hereafter called LAI)

contained 10 mg/ml protein after dialysis against PBS to

remove ascorbate. LAI was stored in 0.5 ml aliquots at -20 C

for future use.




- 15 -


Binding to Rhizobium Cells

Cells washed in PBS were tested for evidence of binding

with LAI by immunofluorescence, tube agglutination, and

slide agglutination. Since several Rhizobium strains

flocculate spontaneously due to cellulose microfibril

production (38), uniform cell suspensions were obtained

by filtering cells in PBS through glass wool loosely packed

in Pasteur pipettes (5 x 10 mm). Cell suspensions were

added to equal volumes of LAI, incubated at 30 C (tubes)

or 22 C (slides) for 4 h, and then examined for agglutination.

In other studies, cells were mixed with LAI, incubated

for 4 h, and washed twice with PBS in the centrifuge at

1,400 x g. Antigenic material in LAI capable of firmly

binding to cells (not removed by washing) was detected

by agglutination and indirect immunofluorescence using

rabbit anti-LAI. This antiserum was prepared according

to the immunization schedule for clover root antiserum and

by itself was unreactive with Rhizobium cells using tube

agglutination and immunofluorescence.

Characterization of Agglutination Factor

Dialysis. The agglutinating titer of LAI using R.

ti6fotii strains 2S-2 and 403 was determined. An aliquot

(3 ml) of undiluted LAI was dialyzed against PBS (3 changes

in a total of 15 liters) and then titrated.

Ultracentrifugation. The agglutinating titer of LAI

before and after ultracentrifugation at 104,000 x g for 1 h

was determined.




- 16 -


Heat liability Aliquots of LA1 were heated at 56 C or

80 C for 10 min, filtered, and titrated.

Enzymatic digestion. Aliquots of LA1 were diluted 1:8

(endpoint agglutination titer) by adding the various enzymes

and other reactive materials in PBS as described above. These

digestion mixtures were incubated for 24 h at 30 C, centrifuged

at 3,000 x g for 10 min, and the supernatants were tested

for agglutination of R.tAifotii strain 403 cells. Strain

403 encapsulated cells were also incubated with the enzymes

for 24 h at 30 C, washed twice by centrifugation at 3,000 x g

for 10 min, and tested for agglutination by LA1 diluted 1:8

with PBS.

Sugar binding specificity. Inhibition of LAl-mediated

cell agglutination by various carbohydrates (Nutritional

Biochemicals Co.) was examined (Table 8). LA1 (0.2 ml)

was diluted 1:8 with the various carbohydrates (final con-

centration 30 mM) dissolved in 0.15 M saline. After 1 h

incubation at 30 C, R. ttLiolii cells (strains 2S-2 and 403)

suspended in PBS were added. These suspensions were

incubated for 2 h at 30 C, and then examined for agglutination.

Filter sterilized soil extract (26) and strawberry clover

root exudate (33) were also examined for inhibition of

agglutination by LA1. Haemagglutination inhibition studies

using LA1 and anti-white clover root antiserum preincubated

with various sugars were 'performed on washed rabbit

erythrocytes adsorbed with 403-AC according to Nowotny (40).

Inhibition of Infection of T. epenz by 2-Deoxyglucose

Fahraeus slide assemblies (20) of T. repens seedlings




- 17 -


inoculated with 2.6 x 108 R. tAi6olii strain 403 cells were

prepared with various increments of sterile 2-deoxyglucose

(final concentrations of 0, 2.5, 5, 10, 30, 50, 100, and 200

mM) or a-D glucose (200 mM) in duplicate. Slides were

incubated for 20 days in a plant growth chamber (Warren

Sherer, Model CEL 255-6, Marshall, Mich.) programmed at 22 C

isothermal, 12 h photoperiod, 18.6 lux light intensity, and

then examined for evidence of root hair infection by phase

contrast microscopy.

Adsorption of R. tifotfii Cells to Root Hairs of T. teens

Slide assemblies of T. Aepens seedlings without agar

were inoculated with approximately 1.9 x 10 infective or

noninfective R. txidfoii cells. After 12 h incubation in

the growth chamber (10 h darkness followed by 2 h light), the

slides were removed. The cover slips were removed, then the

roots were washed on the slides by a gentle stream of saline.

The slide assemblies were filled with saline, new cover

slips were added, and then the roots were examined by phase

contrast microscopy. Four optical median planes of two

primary roots per strain were evaluated. The number of

bacterial cells per strain adsorbed to root hairs approximately

200 pm in length was determined and the data were evaluated

statistically.

Electron microscopic examination of R. thifolii associated

with T. fAugi6enum root hairs was performed as previously

described (38).













RESULTS AND DISCUSSION

Antigenic Cross Reactivity between R. tVitolii and Clover Roots

Tube Agglutination

Antiserum to white clover roots cross agglutinated both

infective and noninfective R. tAiLotii strains (Table 2).

Endpoint cross agglutination titers were higher for the

infective strains in 3 out of 4 infective-noninfective strain

combinations. In the fourth combination, both the infective

and noninfective strains had equal agglutination titers. Cells

were not agglutinated in normal serum controls. These results

indicated that cross reactive antigens between clover roots

and the surface of R. ttifolii strains exist, and that the

surface antigenic composition of the infective strains may be

different from the noninfective strains.

Immunofluorescence using Anti-Clover Root Antiserum

Unadsorbed anti-white clover root antiserum was reactive

with the surfaces of infective and noninfective R. ttifotii

cells when examined by indirect immunfluorescence (Table 3,

Fig. 1). The anti-white clover root antiserum was adsorbed

with cells of each combination of infective and noninfective

strains of R. tL6otNii. When adsorbed with whole cells of

the noninfective strains, this antiserum remained reactive

only with infective cells, including the three spontaneous

infective revertants 0435-21, BA-L and BA-S (Table 3). These

three revertants also reacted strongly with antiserum against


- 18 -




- 19 -


Table 2. Agglutination o6 Rhizobium ti6oZlii with
Rabbit Anti-iATrzotium Aepeins
Root Antcsevum


Strain



2S-2a

2Lb

403a

Bart Ab

T37a

Bio-9b

0435a

0435-2b


aInfective on T.

boninfective on
Noninfective on


Endpoint titer


epenas.

T. Aepens.




- 20 -


Table 3. Reactivity of Anti-TAiotZium Aepen. Root Antidseum
and Rhizobium ftidolii Ceilu Using
Ind-irect Immunofluotescence


Antiserum

Adsorbed with Adsorbed with
Strain Unadsorbed noninfective cells infective cells

2S-2a + +

2Lb +

403a + +

Bart Ab +

0435a + +

0435-2b +

T37a + +

Bio-9b +

WU290-Ia + +

WU290-Nb +

0435-2Ia + +

BA-La + +

BA-Sa + +

Jla + NDc ND

J2a + ND ND

NA30a,d + ND ND

NA-30a,e + ND ND


Infective on T. Aepens.

bNoninfective on T. Aepens.

ND, not determined.

Grown on a chemically defined medium.
eGr in soil extract.
Grown in soil extract.




































Fig. 1. Immunofluorescence of Rhizobium tri6otii
403 (infective) using anti-TLiaoZium repens root antiserum.
Bar scale equals 2 ym.




-22-


LIl


i




- 23 -


the parent infective strains. When adsorbed with cells of the

infective strains, the root antiserum was reactive with

none of the R. t ifolii strains examined. These results

indicated that R.' tttAioii cells have antigens on their

surfaces which are cross reactive with clover roots, and

there is a distinctly greater degree of antigenic cross

reactivity displayed by the infective strains. A portion of

this homologyy" is lost when cells lose the ability to

infect the root hairs of their host, but it is reacquired

when the cells spontaneously revert back to their infective

state. R. tAifoZi strains J1 and J2, which were recently

isolated from root nodules of natural stands of white clover,

were also reactive with the anti-white clover root antiserum.

This indicated that the antigenic cross reactivity of

R. tai-oii and clover is not unique only to laboratory strains

maintained for years as stock cultures. While growing in

soil extract, R. tlijolii NA-30 (infective) was capable of

maintaining a cell surface which was antigenically cross

reactive withwhite clover roots. Controls for autofluorescence

of cells and non-specific staining of FITC-labeled anti-rabbit

y globulin were negative (no fluorescence). Preimmune sera

were unreactive.

Anti-white clover root antiserum was tested for reactivity

with a variety of other Rhizobium species, including R. japonicum,

R. ZeguminosaAum, R. phabeoti, R. meli-oti, and members of the

so-called cowpeaa miscellany,"none of which infected clover

(Table 1). None of these Rhizobium species were reactive with

antiserum to T. Aepens immunofluorescencee) except R. sp. strain




- 24 -


HR1. In this latter case, the preimmune normal serum control

was equally reactive, and therefore the rabbit had natural

antibody reactive with surface antigens of this organism.

These results indicate that strains of Rhizobium incapable of

infecting clover lack specific surface components which are

antigenically cross reactive with clover roots.

Antisera were prepared against 27,000 x g supernatant

and 104,000 x g pellet fractions of macerated root antigens

of strawberry clover. Antisera to both fractions reacted

with infective strains of R. tAifolii (2S-2, 403, T37, and

0435). This indicated that R. thidotii was antigenically

cross reactive with at least 2 host species in the clover

cross-inoculation group. It seems feasible to predict similar

results with other clovers that can establish successful

symbiotic relationships with R. trLA6oii. Cross reactive

strawberry clover root antigen(s) remained in the supernatant

at 27,000 x g for 1 h, and were sedimented at 104,000 x g

for 3 h in sufficient quantity to elicit an immune response.

A higher degree of antigenic similarity exists among legumes

in the clover cross-inoculation groups (1).

h dioimmunoassay

Radioimmunoassay was performed to detect and quantitate

the degree of antigenic cross reactivity between R. Ai6otii

strains and the roots of their clover host. The amount of

cross reactive antibody bound to equal numbers of strain 403

(infective) and Bart A (noninfective) cells after repeated

additions is shown in Fig. 2. At saturation, approximately

twice as much cross reactive antibody bound to the infective




- 25 -


S403





0








2 3 4
















and Bart A (noninfective) cells. Cells of both strains were
adjusted to equal population sizes following direct micro-
scopic counting using a Petroff-Hausser chamber. Cells were
incubated with rabbit anti-T. epens root antiserum, washed,
and then incubated with I-labeled goat anti-rabbit y
globulin. After each addition, cells were incubated, washed,
and counted for y emissions. Points and bars represent means
and standard deviation.




- 26 -


as compared with the noninfective cells. These data confirmed

the earlier observations (tube agglutination and immuno-

fluorescence) that quantitative differences in the degree of

antigenic cross reactivity exist between clover roots and

infective strains of R. ttiZoZii as compared with noninfective

strains.

Immunofluorescence using Anti-R. taif4o-ii Antiserum

Axenically grown white and strawberry clover seedling

roots were reacted with antisera prepared against several

infective R. tLiotlii strains (2S-2, 403, T37, and 0435),

and then examined for the presence of bound cross reactive

antibody using the indirect immunofluorescent technique.

Antibody to each of the infective strains bound to the

surface of sterile clover roots. Immunofluorescence by

anti-0435 antibody appeared brightest at the growing root

hair tip of white clover as compared with other external

root parts. In all cases, immunofluorescence was present on

the exposed surface of the root epidermal cells including

root hairs (Fig. 3) and diminished at the root meristem.

These results indicated that clover has antigens on its

exposed root surface which are cross reactive with antigens

from infective R. tLi6otii cells.

The blue autofluorescence of fresh root tissue (27) was

removed by a K490 barrier filter. Jones and Russell(27) used

immunofluorescence to identify R. tAidolii strains in T. Aepens

nodules. Interestingly, their photographs provided clear

evidence of antibody to R. LAi6otii binding to surface

antigens of clover roots without the authors actually reporting

the observation.
































Fig. 3. Immunofluorescence of a Ttifotium Aepend
root using anti-Rhizobium tti-olii 2S-2 antiserum. Bar
scale equals 45 pm.




-28-




- 29 -


Antisera to the infective strains of R. tui6olii

previously adsorbed with whole cells of the corresponding

noninfective strains bound to the antigens of the white

clover root surface and the homologous infective R. thLiaoii

strains. Antibody to R. tAijotii 403 (infective) which was

cross reactive with the surface of clover roots could be

removed by adsorption with encapsulated 403 cells. These

results indicated important differences in the degree of

antigenic cross reactivity between infective vs. noninfective

R. tLofioti strains and the roots of a clover host.

Analytical Studies of Rhizobium Capsular Antigens

Since the cross reactive antigen of R. tri6ooii 403 was

localized on a morphologically distinct capsule as viewed by

immunofluorescence, the capsular material was isolated and

characterized.

Precipitation and Immunofluorescence

When dissolved in water, 403-AC was optically clear

(Fig. 4). The 403-AC and the capsular material from Bart A

(BAC) completed with soluble DEAE-dextran (Fig. 5), cetyl-

pyridinium chloride, hexadecyltrimethyl ammonium bromide, and

acian blue. These compounds form specific precipitates

with acidic polysaccharides (11, 19, 35, 40). 403-AC and BAC

did not react with dextran sulfate which precipitates basic

polysaccharides (19). The 403-AC also formed a precipitate

when reacted with anti-white clover root antiserum (Fig. 6)

and a clover seed extract (Fig. 7, discussed later). The

quantitative precipitin curve (Fig. 8) indicated that 403-AC

was cross reactive with white clover roots. A maximum of


























Fig. 4. Rhizobium tLi6otii 403 cap'sular antigen
dissolved in water (2 mg/ml).

Fig. 5. Reaction of Rhizobium tti6otii 403 capsular
antigen with DEAE-dextran.

Fig. 6. Reaction of Rhizobium tti6otii 403 capsular
antigen with anti-T/iotium Aepens root antiserum.

Fig. 7. Reaction of Rhizobium tifdolii 403 capsular
antigen with Thifotium Aepend seed extract.




-31-


-


- .- --


C __
k"


-


I~"~F~"_'lr~E~_~_~~-


II~


,. ...-,




- 32 -


Atipi A.LWad Crig)


Fig. 8. Quantitative precipitin curve of Rhizobium
trifoZii 403 capsular antigen with anti-TLifotiunm Lepens root
antiserum. Symbols at the top represent excess antigen or
antibody in the supernatants after removal of the imniuno-
precipitates by centrifugation.




- 33 -


930 pg cross reactive anti-white clover root antibody protein

was present per ml of undiluted antiserum. Antibody against

white clover roots and R. tifdotii 403 reacted with 403-AC

using immunofluorescence.

Electrophoresis

At pH 8.6, 403-AC and BAC were electrophoretically

homogeneous. They both had a net negative charge, and

migrated as single bands of equal electrophoretic mobilities

(2.08 x 10-5 cm2/V sec).

Gel Permeation Chromatography

The capsular antigen 403-AC eluted as a sharp peak

just after the void volume of the glass bead columns of AX

through FX (Waters Assoc.) and Bio-Glas 2500 (Bio-Rad, Fig. 9).

No additional peaks of lower molecular weight in 403-AC were

eluted within the selective permeation range of the beads.

When compared with the elution volumes of various markers,

these results indicated that 403-AC had an average molecular

weight in excess of 4.6 x 106 daltons. This material was not

contaminated with detectable amounts of any smaller compounds

that would have changed the refractive index of water.

biochemical Component Analysis

The capsular antigens 403-AC and BAC consisted primarily

of carbohydrate. Total neutral sugar (anthrone) and uronic

acid (carbazole) contents of 403-AC were 68.0 + 0.5% and

32.0 + 0.5%, respectively. Gas-liquid chromatography of the

trimethylsilylated (TMS) derivatives of the sugars released

by acid hydrolysis of 403-AC indicated 2-deoxyglucose,




- 34 -


Bio-Glas 2500
HPLC-Gel Permeation of 403AC

Vo


Fig. 9. High pressure liquid chromatography of the
Rhizobium trifolii 403 capsular antigen on Bio-Glas 2500.
Tobacco mosaic virus (TMV), Bromegrass mosaic virus (BMV),
and dextran markers are 3 x 107, 4.6 x 106, and 2 x 106 dal-
tons, respectively.




- 35 -


a-D galactose, a and $-D glucose, e-D glucuronic acid, and one

unidentified compound (Fig. 10). The unidentified compound

(peak 2) did not correspond to the trimethylsilylated

derivative of L-sorbose, L-fucose, 2-deoxyribose, 2-deoxy-

galactose, 3-0-methyl D-glucose, D-ribose, 6-deoxy-l-galactose,

D-mannose, D-fructose, L-arabinose, D-xylose, 4-O-methyl

D-glucuronic acid, D-melibiose, or L-rhamnose. Hexa-

methyldisiloxane was also present in the mixture as a

normal by-product of the silylation reaction (43). BAC

gave an identical chromatogram (not shown). The absence of

deoxyribose and ribose indicated a lack of' nucleic acid

contamination. The analytical system employed could detect

1 nanogram of a-D glucose TMS as a lower limit of sensitivity.

Trace amounts (1-10 nanomoles per mg dry wt) of

lysine, aspartic acid, threonine, serine, glutamic acid,

glycine, alanine, and isoleucine were identified in acid

hydrolysates of 403-AC by comparison with 44nm ratios and
570nm
identical retention times of authentic standards. The

total amino acid content could account for approximately

0.26% of the dry weight of 403-AC and therefore was not

considered a major structural component of the capsule.

Neither glucosamine nor galactosamine was detected within

the sensitivity of the amino acid analyzer employed (10

nanomoles).

Diffraction and Spectral Analyses

X-ray diffractograms of 403-AC and BAC revealed no

reinforcement peaks within the range 2* 20 to 600 20. This




- 36 -


OLC of 403 AC HYDROLYSATE TMS


RETENTION---


Fig. 10. Gas-liquid chromatography of trinethylsi-
Ivlated sugars in an acid hydrolysate of Rhizobium triioiii
403 capsular antigen. Peak 1= pyridine solvent, peak 2= un-
identified, peak 3= 2-deoxyglucose, peak 4= a-D galactose,
peak 5= a-D glucose, peak 6= hexadisiloxane, peak 7= B-D
glucose, peak 8= a-D glucuronic acid.




- 37 -


indicated an amorphous, non-crystalline structure, and the

lack of detectable contamination with cellulose microfibrils

which were produced by both strains (38). The infrared

spectra of both materials were essentially identical

(Fig. 11). The spectra were consistent with a carbohydrate

structure rich in hydroxyl and carboxyl groups. The
-i
absorption band at 890-900 cm- was characteristic of a
-i
6 glucosidic linkage and absence of a band at 870 cm-

indicated absence of an a glucosidic linkage (2). The UV

spectrum of 403-AC indicated absorption in the short UV range

only (210-230 nm). Lack of absorption down to 230 nm

indicated no detectable nucleic acid and/or protein which

would absorb in this range.

It is concluded from these analytical studies that 403-

AC is a water soluble, amorphous, high molecular weight,

$-linked, acidic heteropolysaccharide.

Antigenic Cross Reactivity of R. traiolii Strain 403 After
Enzymatic Digestion

Encapsulated R. tA'iooii 403 cells were treated with

various enzymes and other reactive materials and then tested

for cross reactivity with anti-white clover root antiserum

using immunofluorescence. The antigenic cross reactivity of

strain 403 capsules was eliminated by lysozyme, periodate,

acid (HC1, pH 3) and alkaline (NaOH, pH 12) treatments.

Lysozyme treatment lysed the cells. Apparently the capsular

antigen was washed away from the cell walls solubilized by

lysozyme. The capsules of the cells were removed by washing

after periodate treatment. However, the underlying cell




- 38 -


INFRARED SPECTRA


Coao oo o0o ao woo woo MOO nmO mo o
WAVENUMBER (cm-1)






Fig. 11. Infrared spectra of Rhizobium th.i6oii
403 and Bart A capsular antigens.




- 39 -


surface remained antigenically cross reactive with clover

roots. This indicated that these cells maintained an

antigenically cross reactive surface even if they lost their

capsules. Maintaining an underlying cross reactive layer

could possibly be important for the cell in the rhizosphere

if the C/N ratio of root exudates restricted abundant

capsule formation.

Effect of pH on Antigenic Cross Reactivity

The antigenic cross reactivity of R. tAliolii 403 cells

following acid and alkaline treatment was examined in more

detail. The degree of antigenic cross reactivity remained

high in the pH range 6-8, but diminished outside of this

range (Fig. 12). The acid sensitivity of the cross reactive

surface antigen (at pH 5 and below) corresponded to the

acid-sensitive step where root hair curling, infection, and

nodulation abort (36). If possession of the cross reactive

antigen is essential to the infection process as proposed

here, then its loss at high H+ concentration may provide

a biochemical explanation of nodulation failure in acid soils

despite proper inoculation.

Ecological Activity of R. tifotii Capsular Antigens

Rhizobium extracellular nondialyzable material has been

known to induce deformation of legume root hairs (23, 30, 33,

54). A markedly curled deformation of root hairs has been

reported to be restricted to inoculation of a legume with

live, virulent, homologous Rhizobuum cells (54). Other have

shown that root hair deforming substances produced by




- 40 -


5
-HCI


8 9 10 11
pH Na0OH--


Fig. 12. The effect of pH on the immunofluorescent
cross reactivity of encapsulated cells of Rhizobium VtLifoii
403 with anti-Tli6otium repens root antiserum. The K470
barrier filter was used to differentiate reactive and un-
reactive cells.




- 41 -


R. tci6olii could be adsorbed to white clover roots and

eluted again by acetic acid or urea (46).

The antigenically cross reactive material, which was

isolated by the method of Nowotny (40) from infective and

noninfective cells grown on a chemically defined medium,

was tested for its ability to deform root hairs of straw-

berry clover grown axenically. The results are reported in
2
Table 4. A X test indicated a highly significant dependence

(at 99.9% level) of root hair deformation on the addition

of this capsular material. Another X test indicated (at

99% level) that the capsular material from'all the strains

examined did not induce root hair deformation to the same

degree. Several X2 tests were performed to determine if the

degree of deformation was independent of whether or not the

curling factor came from the infective or the corresponding

noninfective strain. These tests were all rejected at the

99% level, indicating that the degree of root hair deforma-

tion induced by the curling factor was significantly

dependent on the strain from which it was isolated. One-

tailed Z tests were performed to test if the frequency of

observations within a deformation category (straight,

branched, moderately or markedly curled) was greater for one

strain than for another (e.g., infective vs. noninfective).

The frequency of straight root hairs was significantly less

(at 99% level) when incubated with the capsular material of

the infective strains as compared with the corresponding

noninfective strains. In all cases except the branching





- 42 -


WD N 'a' uo Lo mcm r-4 Hn
m D N O Co M Hr-i
m N N N H IQ N mi











Cd (C 4-4 a)
Soo o < o eo

N C C C 0 C)
m N m


9
C'






+4
0
0





cEd






Ns
to .
Cd Cdd-











Cd 9
Cd+
Cdi U^







o C0

U
Ocd C

YO C

Ed











Cd
.0 Q
Cd -


0


in










T m

m n
O-4










0)
co

01


'rc
NC
U I

SCm p-
o o E-


Cd Cd
m N 0)a'*C

NN N C- C
N] -a









Cd
a' '. LCnCCO C

o CI) N C
N










Cd Cd
N H C') CC

cI) 0 Cl)m C
ICCco H 0


rou

0 'Cd
. Nc l ,


-J
r-1








4 -













0.
a '







Q- -m




: 0 4-
-r-

U in









041





S0
H -H *


Q,

- $ 0a 3















H3 0 4-
SC 0n



- C -1-

x a) z








4-4 0 O ) Cd
S '9 0H
4 o Cd E C

0 0(d H C

$4 I- 0
rd C
4a 0









C. m z H C

0 rd > U. aH u









rd V 4-4 M 4-4
m U d -4H 4-


H) rd 0 41

.0 u C 4-4
H0 Cm (












0 -t1 **I () -' -
da alC 41' U -


cn n

m












H o











M r-

0




- 43 -


frequency of 2S-2 and 2L, the degree of root hair deformation

was significantly greater for the infective strains. Thus,

the antigenically cross reactive capsular material of

these R. tiSotlii strains possessed biological activity

characterized by the ability to deform root hairs of the

clover host. Photomicrographs of control root hairs and

root hairs deformed by 403-AC and BAC are presented in

Figs.13, 14, and 15, respectively.

Clover Lectin Studies

A clover lectin capable of differentiating infective

and noninfective R. tAi6olii strains was sought. A heat

sensitive, nondialyzable, soluble protein or glycoprotein

agglutinating factor with definable sugar specificity has

been found in LAI. The agglutinating factor reacted with

the cross reactive surface antigens which these symbionts

share.

Binding to Rhizobium Cells

The spectrum of cell agglutination by LAI (Table 5)

included all R. tifaotii strains capable of infecting

white clover (including the infective revertants) but

none of the noninfective R. tLifolii strains or other

Rhizobium species incapable of infecting this legume. R.

LeguminosaAum 3HOQ1 and R. sp. HR1 could not be examined

due to autoagglutination. Control and LAI-agglutinated

cell suspensions of R. tAi-otii 2S-2 are illustrated in

Figs.16 and 17, respectively. The ability to agglutinate

indicated that the proposed clover lectin was at least

divalent, i.e., had at least two reactive binding sites.

































Fig. 13. TAidolium ftagifetum root hairs.
Bar scale equals 45 pm.

Fig. 14. TAi6olium 6(Aagie6um root hair de-
formation in the presence of the capsular material
from Rhizobium hri6olii 403 (100 pg/ml). Bar scale
equals 45 pm.

Fig. 15. TALiotium fragierLum root hair de-
formation in the presence of the capsular material
from Rhizobium tri6olii Bart A (100 pg/ml). Bar scale
equals 45 pm.




-45-


7


7 .1- *
rat>- L-s. -$s-.-fc _* ^f .,
r 1 -- -- --T
h-.--------'-r" r-,4--


S**1




- 46 -


I


O H Q HQ
co o) o
N O O O N N N HN
m N oomoN u H


E1




2i
I-

o












Lu
-9






N
E -u







..9e




an



0
.0 &
L <


O










0

tH
S:





'
**<
*^
'3








r-l
a
(0


.9

0

-d 0
0


+I+I+ I + + + + ++


.0

0
4Jtf


.0

OrrJ rrJl T
I i C/)
00 0 I I
-H H N F <
CO a 1 l-


I


.Q
o.)
to ,-l
o O0
.-o 0
- m
m mr~


E



3 0


a 9
0
o d

Q- CT
*. .1

CX' C^


.9
0:
.9

0~


N 10

cqO
C13 q :J


.9
.9

0


*




4 o


-rI



.- 0

O> 0J
-H 0c


4-) 4 0
rl







0 -
-H ZD u


rU C

. a %































Fig. 16. Rhizobium t&i6otii 2S-2 cells suspended in
saline. Bar scale equals 0.5 mm.

Fig. 17. Agglutination of Rhizobium tAifotii 2S-2
cells by the TAidotium Aepens seed extract. Bar scale equals
0.5 mm.





-48-





- 49 -


Whether the agglutinating factor could bind to cells at

levels below the sensitivity of the agglutination test

was then tested. Cells were incubated with LAI, washed,

and then reacted with anti-LAI using tube agglutination

and indirect immunofluorescence. Antigenic material in LAI

capable of firmly binding to the cells was detected on

both infective and noninfective R. tLidotii strains but not

on other Rhizobiumn species incapable of infecting white

clover (Table 6). Indirect immunofluorescence showed that

the proposed clover lectin bound to the distinct capsules

of R. trLiolii strain 403.

Characterization of the Agglutination Factor

The endpoint agglutinating titer of LAI did not change

following extensive dialysis (3,500 molecular weight cut-

off) and ultracentrifugation at 104,000 x g for 1 h.

Treatment of LAI at 56 C and 80 C for 10 min completely

destroyed its ability to agglutinate infective strains of

R. t'Zfotii. These results indicated that the agglutinating

factor was nondialyzable, nonparticulate (soluble), and heat

labile.

The effect of various enzymes and other reactive

materials on the agglutination of R. tifaolii 403 cells by

LAI was examined. Pronase, trypsin, periodate, acid (HC1,

pH 3), alkaline (NaOH, pH 12), and 7 M urea treatments of

LAI destroyed its ability to agglutinate the cells (Table

7). This indicated that the agglutination factor (lectin)

in LAI was an acid and alkali sensitive protein (or





- 50 -


cd
u
Cd








Cr
0



































(A
I0-r


-4 1




















Cd








0
Cd .-

i-rC
I -
-4.4
Cd


'CN
CI .0r

(N (N


U 0 u
K4 U cJ 0>0C
F~~ HO UO
m Lfl I AD O-I U U

rc- 0 rn () H C, 0 0 mo
(Y) 'r z r- H ) 0 ) 0 -4
p C) 0 m~ (n Nl Nl N


Cd
0
Cd Ce C
Cd Cdi


I I


+ +.+ + + + + + I I


























+ + + + + + + + 1 I


I I


0 U
Hl 0



N H-l 0
Uhlrl


s
n
'd
a:
c!


.-
0
Cd -~
-4
4JI








>d U
-0 44
Cd Ce
Cd Cd



Cd -






,-4 44
U Cd



l r: 0
Cd. Ur
Iq H 2
co wQ


I I




- 51 -


Table 7. E{ject o6 Various TAeatments on Aggtutination
of Rhizobium Vtifolii 403 by
Clovei Seed ExtLact


LAla Cells
Treatment treated treated


Pronase +
Trypsin +
Deoxyribonuclease + +
Ribonuclease + +
Phospholipase + +
Cellulase + +
Pectinase + +
Periodate
HC1 (pH 3)
NaOH (pH 12)
7 M Urea +
1% Triton X-100 + +
Soil extract + +
Clover root + +
exudateb
Untreatedc + +


aLA1, T. Repens seed extract.

bT. fLagiferum.
CEnzyme buffers alone.




- 52 -


possibly glycoprotein) which may require the participation

of H-bonding for agglutinability. These experiments did

not distinguish between the possible need of H-bonding for

intramolecular stability of the lectin and for intermolecular

binding of the lectin to the surface of the RhizobiuLm cells.

Periodate, acid, and alkaline treatments rendered the 403

cells non-agglutinable by LAI as with the anti-white clover

root antiserum as described earlier. Clover root exudate and

soil extract did not prevent the agglutination of R. tAiotii

403 cells by the clover lectin. R. t.i6oZii NA-30 was also

agglutinated by clover lectin after growth for one generation

in soil extract.

Sugar Binding Specificity

Inhibition of LAI-mediated agglutination of R. tLifolii

403 and 2S-2 cells by various carbohydrates was examined.

The sugar inhibition patterns were identical for both

infective strains. Preincubation of LAI with 30 mM

2-deoxyglucose or N-acetylgalactosamine inhibited its ability

to agglutinate the cells (Table 8). None of the other

carbohydrates at this concentration were inhibitory.

LAI agglutinated rabbit erythrocytes coated with

403-AC only if followed by anti-LAI. This antiserum did

not agglutinate cells coated only with 403-AC. LAI-mediated

passive haemagglutination was inhibited by 30 mM 2-deoxy-

glucose and N-acetylgalactosamine (Table 8). Identical

results were obtained with anti-white clover root antiserum.

Since 2-deoxyglucose was a component of this cross reactive




- 53 -


Qfu



0


>1P



r. 0
C4r'

.,-
0.Q0
















MO




ud=




pe
ut


* a?


'0
S-4 0


0 0 0 0
do o



00 00 00
S.)- O nd o0

O ( 02 0 3 I 0 0

0 1 0 r4 40 l O 0 O HU > i
00000020-2 O H0
0,-0 00 00 0 H 0 ( '
Q I I 0 H I 4 00
I I I I 0 Q I I I I I I I
0)0)00010 0)C.4


+ + + ++++++++ I + +


'9








0

40
0)








m 1
S 0










0d
4-)
m















0 0
0) 00
















tp 4)
UO 4-
0 0U'










+ w
0 0
4. 0 .1-














0 21
0 0 >

4-, 4 '0





0 0 a0
c m 1
O O 0 'd








0 -H O w
H O4Q
40 0 -l
' 0















U) j d C)
a< EM a4 ;2
o Qo o
10 0 n0




- 54 -


capsular antigen, the sugar binding site of both the clover

lectin and the anti-white clover root antibody was

probably directed toward this sugar.

Inhibition of Infection of T. /epegn by 2-Deoxyglucose

Since 2-deoxyglucose inhibited the binding of the clover

lectin to the cross reactive antigen of R. tifolii, it was

hypothesized that this sugar may interfere with the bacterial

infection of clover by occupying all the lectin binding

sites. Infection of white clover root hairs by R. tAidolii

403 was completely inhibited by 2-deoxyglucose at concentra-

tions of 30 mM and above (Table 9). Partial inhibition was

evident at lower concentrations (2.5-10 mM). The inhibition

was specific for 2-deoxyglucose and was not due to hypertonic

osmotic effects since infection and nodulation occurred in the

presence of 200 mM a-D glucose. However, 2-deoxyglucose is

known to interfere with glucan wall synthesis, microfibril

production, and cell metabolism in Sacchauomyced cexeviziae

(32), and therefore the inhibitory effect should be regarded

at the present time as a result of multiple effects.

Adsorption of R. thi6olii to Root Hairs of T. Aepen6

C mparison of Infective and Noninfective R. thidotii Strains

The mean number of R. tvikolii cells adsorbed to root

hairs of approximately 200 pm in length after 12 h incubation

are presented in Table 10. When pooled together, the mean

number of adsorbed infective and noninfective cells were

25.8 + 5.9 and 5.6 + 1.9, respectively. The populations

were compared statistically using the t tests computed after




- 55 -


Table 9. Inhibition o6 Indection and Nodulation o6
Tti6otium trepens with Rhizobium
-triAo.-ii 4. 4 y 2-deoxyguc.ose


Concentration of Infection threads Nodules on
2-Deoxyglucose on two roots two roots
(mM)

0.0 25 12

0.0a 23 9

2.5 6 0

5.0 3 0

10.0 1 0

30.0 0 0

50.0 0 0

100.0 0 0

200.0 0 0


aContaining 200 mM a-D glucose.




- 56 -


Table 10. AddoIption o6 Rhizobium t'tiotiL CeftM
on Ti 6 oZiZtm ei pans R oo-t Hafcwt


Adsorbed cells
(means + std. dev.)


21.00d + 1.00

3.00 + 1.73

22.50d + 2.81

6:67 + 1.63

22.75d + 2.22

4.80 + 1.64

25.50d + 4.12

8.00 + 2.45

25.67d + 0.58

5.33 + 2.31

37.33d + 9.48


aOn root hairs of approximately 200 pm in length.

Infective on T. Aepens.

CNoninfective on T. Aepena.

dSignificantly greater than corresponding noninfective
mutant (at 99.5% level).


Strain


2S-2b

2Lc

0435b

0435-2C

T37b

Bio-9c

WU-290-Ib

WU-290-NC

403b

Bart Ac

BA-Lb




- 57 -


the square root transformation (47) of the means. For

every strain combination examined, the mean number of

infective cells adsorbed was significantly greater (at 99.5%

level) than the noninfective cells. Of particular importance

was the finding that the infective revertant BA-L had a

significantly greater frequency of adsorbed cells than the

corresponding noninfective strain Bart A. Thus, it can be

concluded that the phenotypic trait of infectivity for

R. thifolii is directly correlated with the extent of cell

adsorption to root hairs of the clover host. Another study

has indicated no correlation between the extent of root

adsorption with the infective capabilities of the

microorganisms (42). This study was based on indirect

measurements involving viable colony counts of cells following

their removal from roots by shaking. Unfortunately this

latter approach did not distinguish root hairs from other

adsorptive root surfaces, nondispersable flocs from single

cells, and the variability of root surface areas among

individual plants.

Adsorption of the rod-shaped bacteria occurred

primarily through attachment of one of their poles to the

root hair surfaces (Fig. 18). Polar orientation has been

reported as the position of attachment of single cells of

R. trAlolii (38), R. meitloti (12), R. japonicum (44), and

R. sp. (Ae.chynomene nodulating strain, [37]) on their

appropriate legume host root surface.

Ultrastructure of Adsorption

The ultrastructural details of the association of































Fig. 18. Adsorption of Rhizobium tAiotZii 403 cells
on a Tri6joium repens root hair. Cells are polarly attached.
Bar scale equals 20 pm.

Fig. 19. Electron photomicrograph of an ultrathin
section of Rhizobium t/Liotii NA-30 in association with a
TiLioLium (ragifeAum root hair. The fibrillar capsule (A)
of the bacterium is in contact with electron dense particles
(B) on the outer periphery of the root hair cell wall (C).
Bar scale equals 1 pm.




-59-


IrI








B




C -i





r~i ''19




- 60 -


R. tif6olii NA-30 with the clover root hair wall are shown

in Fig. 19. This photograph was typical of the polar

orientation that was routinely observed. The fibrillar

capsule of the bacterium was in physical contact with

electron dense, globular aggregates lying on the exterior

periphery of the fibrillar root hair cell wall.

Based on the data of these investigations, the

following model was proposed (Fig. 20) to explain this

preinfective adsorption event which contributes to host

specificity in the Rhizobium-clover association. Infective

strains of R. tAiotlii have on their surfaces a polysaccharide

which is antigenically cross reactive with an antigen on the

root surface of the clover host, and also is capable of

deforming root hairs. It is proposed that the clover lectin

which recognizes these surface antigens cross-bridges them

to form a correct molecular interfacial structure which

allows for specific adhesion of the bacteria to the root

surface. Following this specific cell adhesion, the

invagination process of the root hair wall begins, resulting

in infection thread formation (39). Noninfective R. tAiZotZi

cells have cross reactive surface antigens but either in

reduced quantity or sterically blocked as neither antibody

nor lectin binds to them as efficiently as to infective cells.

Other Rhizobium species lack this surface cross reactive

antigen, and therefore do not bind to the clover lectin.

The model is consistent with the findings of Hamblin and

Kent (22), and Bohlool and Schmidt (4) who introduced the

role of lectin in the adsorption of the Rhizobium. But the




- 61 -


Lair wall


Root hair I I
cyteplasrm Ip--Af




il Crss- rectivw ntien







Fig. 20. Schematic diagram of the proposed cross-
bridging of the cross reactive antigens of Rhizobiumi tai6oZii
and Tti6oZium root hair with a clover lectin.




62 -



lectin alone does not confer specificity. It is one component

of a multimembered specificity-determining complex. Two

other components of this complex are the cross reactive

surface antigens.













LITERATURE CITED


1. Baldwin, I. L., E. B. Fred, and E. G. Hastings. 1927.
Grouping of legumes according to the biological reactions
of their seed proteins. Bot. Gaz. 83:217-243.

2. Barker, S., E. Bourne, and D. Whiffen. 1956. Use of
infrared analysis in the determination of carbohydrate
structure. Meth. Biochem. Anal. 3:213-245.

3. Belue, G. P., and G. D. McGinnis. 1974. High pressure
liquid chromatography of carbohydrates. J. Chromatogr.
97:25-31.

4. Bohlool, B. B., and E. L. Schmidt. 1974. Lectins: a
possible basis for specificity in the Rhizobium-legume
root nodule symbiosis. Science 185:269-271.

5. Brethauer, T. S., and J. D. Paxton. 1975. Interaction
of soybean lectin with extracellular material from
Rhizobium bacteria. Abstr. North Central Division
Meeting, Amer. Phytopathol. Soc., June 23-24, 1975,
West Lafayette, Ind.

6. Cahill, J. F., B. C. Cole, B. B. Wiley, and J. R. Ward.
1971. Role of biological mimicry in the pathogenesis
of rat arthritis induced by Mycoplasma a/thtlitidis.
Inf. Immun. 3:24-35.

7. Cappelletti, C. 1926. The bacteroid-like form and
immunity in leguminous plants. Internatl. Congr. Plant
Sci., Ithaca, N. Y. Proc. 1:59-60.

8. Chada, K. C., and B. I. Strivastava. 1971. Evidence for
the presence of bacteria-specific proteins in sterile
crown gall tumor tissue. Plant Physiol. 48: 125-129.

9. Charudattan, R., and J. E. DeVay. 1972. Common antigens
among varieties of Gossypiurn hi-sutum and isolation of
Fu6araium and VeAtic.iiCiam species. Phytopathology
62:230-234.

10. Charudattan, R., and D. H. Hubbell. 1973. The presence
and possible significance of cross reactive antigens in
Rhizobiumn-legume associations. Antonie van Leeuwenhoek
J. Microbiol. Serol. 39:619-627.


- 63 -




- 64 -


11. Clapp, C. E., and R. J. Davis. 1970. Properties of
extracellular polysaccharides from Rhizobium. Soil Biol.
Biochem. 2:109-117.

12. Dart, P. J., and F. V. Mercer. 1964. The legume rhizo-
sphere. Arch. Mikrobiol. 47:344-378.

13. Dazzo, F. B., and D. H. Hubbell. 1974. A quantitative
assay of insoluble polyvinylpyrrolidone. Plant Soil
40:435-440.

14. Dazzo, F. B., and D. H. Hubbell. 1975. Antigenic
differences between infective and noninfective strains
of Rhizobium thLiolii. Appl. Microbiol. 30:172-177.

15. Dazzo, F. B., and D. H. Hubbell. 1975. Concanavalin A:
lack of correlation between binding to Rhizobium and
specificity in the Rhizobium-legume symbiosis. Plant
Soil 43:717-722.

16. DeVay, J. E. 1975. Protein specificity in plant disease
development: protein sharing between host and parasite.
NATO Conference on Specificity in PlaAt Diseases, May
4-16, 1975, Sardinia Advanced Study Institute, Sardinia.

17. DeVay, J. E., W. C. Schnathorst, and M. S. Foda. 1966.
Common antigens and host-parasite interactions, p. 313-
328. In C. Mirocha and I. Uritani (eds.), The dynamic
role of molecular constituents in plant-parasitic
interactions. Bruce Pub. Co., St. Paul, Minn.

18. Doubly, J. A., H. H. Flor, and C. O. Clagett. 1960.
Relation of antigens of MeZampsoxa i2ni and Linium
usitatisbimuam to resistance and susceptibility. Science
131:229.

19. Dudman, W. F. 1972. Detection of acidic polysaccharides
in gels by DEAE-dextran. Anal. Biochem. 46:668-673.

20. Fahraeus, G. 1957. The infection of clover root hairs
by nodule bacteria studied by a simple glass slide tech-
nique. J. Gen. Microbiol. 16:374-381.

21. Goldman, M. 1968. Fluorescent antibody methods.
Academic Press, N. Y. 303p.

22. Hamblin, J., and S. P. Kent. 1974. Possible role of
phytohaemagglutinin in Pha.eotlu vulgaviz L. Nature
New Biology 245:28-30.

23. Hubbell, D. H. 1970. Studies on the root hair
"curling factor" of Rhizobium. Bot. Gaz. 131:337-342.

24. Humphrey, B. A. 1959. Occurrence of 4-O-methyl
glucuronic acid in Rhizobium gums. Nature (London)
184:1802.




- 65 -


25. Hunter, W. 1973. Radioimmunoassay. In D. Weir (ed.),
Handbook of experimental immunology. p. 17.1-17.33.
Blackwell Scientific Pub., Oxford, England.

26. Johnson, L. F., and E. A. Curl. 1972. Methods for
research on the ecology of soil-borne plant pathogens.
p. 92-96. Burgess Pub. Co., Minneapolis, Minn.

27. Jones, D. G., and P. E. Russell. 1972. The application
of immunofluorescence techniques to host plant/nodule
bacteria selectivity experiments using TAiuotiuim iepenn.
Soil Biol. Biochem. 4:277-282.

28. Kabat, E. A., and M. M. Mayer. 1961. Experimental
immunochemistry, 2nd ed. Charles Thomas Pub., Spring-
field, Ill. 905p.

29. Li, D., and D. H. Hubbell. 1969. Infection thread
formation as a basis of nodulation specificity in
Rhizobium-strawberry clover associations. Can. J.
Microbiol. 15:1133-1136.

30. Ljunggren, H. 1969. Mechanism and pattern of Rhizobium
invasion into leguminous root hairs. Physiol. Plant.
Suppl. V. 82p.

31. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J.
Randall. 1951. Protein measurement with the Folin
phenol reagent. J. Biol. Chem. 193:265-269.

32. Lratky, A., P. Biely, and S. Bauer. 1975. Mechanism of
2-deoxy-D-glucose inhibition of cell wall polysaccharide
and glycoprotein biosynthesis. Eur. J. Biochem.
54:459-467.

33. MacGregor, A. N., and M. Alexander. 1972. Comparison
of nodulating and non-nodulating strains of Rhizobium
t'ifoli. Plant Soil 36:129-139.

34. Markowitz, A. S., S. H. Armstrong, and D. S. Kushner.
1960. Immunological relationships between the rat
glomerulus and nephritogenic streptococci. Nature
(London) 187:1095-1097.

35. Mowry, R. 1956. Alcian blue techniques for the
histochemical study of acidic carbohydrates. J. Histo-
chem. Cytochem. 4:470.

36. Munns, D. N. 1968. Nodulation of Medicago sativa in
solution culture. I. Acid-sensitive steps. Plant
Soil 28:129-146.

37. Napoli, C., F. Dazzo, and D. Hubbell. 1975. Ultra-
structure of infection and "common antigen"
relationships in Aeschynomene. 5th Australian Legume




- 66 -


Conference, March 18-21, 1975, Brisbane, Australia.

38. Napoli, C. A., F. B. Dazzo, and D. H. Hubbell. 1975.
Production of cellulose microfibrils by Rhizobium.
Appl. Microbiol. 30:123-131.

39. Napoli, C. N., and D. H. Hubbell. 1975. Ultrastructure
of Rhizobium induced infection threads in clover root
hairs. Abstr. Ann. Meeting, Amer. Soc. Microbiol.
April 27 May 2, 1975, New York.

40. Nowotny, A. 1969. Basic exercises in immunochemistry.
Springer-Verlag, N.Y. 197p.

41. Pearlman, P., S. Hammarstrom, R. Lagercrantz, and B. E.
Gustafson. 1965. Antigen from colon of germfree rats
and antibodies in human ulcerative colitis. Annu. N.
Y. Acad. Sci. 124:337-394.

42. Peters, R. J., and M. Alexander. 1966. Effect of
legume exudates on the root nodule bacteria. Soil
Science 102:380-387.

43. Pierce Chemical Company. 1972. Handbook of silyl-
ation. Pierce Chemical Company, Rockford, Ill. 48p.

44. Reporter, M., D. Raveed, and G. Norris. 1975. Binding
of Rhizobium japonicum to cultured soybean root cells:
morphological evidence. Plant Science Letters 5:73-76.

45. Schnathorst, W. C. 1973. Comparative antigenic
structures of virulent and avirulent isolates of
Xanthomnonas mavaceaarum. p. 1330. Proc. 2nd International
Congr. Plant Pathology, Amer. Phytopathol. Soc., Sept.
5-12, 1973, Minneapolis, Minn.

46. Solheim, B., and J. Raa. 1973. Characterization of
substances causing deformation of root hairs of
TiotLium arepens when inoculated with Rhizobium triotii.
J. Gen. Microbiol. 77:241-247.

47. Steel, R. G. D., and J. H. Torrie. 1960. Principles
and procedures of statistics. McGraw-Hill Book Co.,
Inc., New York. 481p.

48. Stevens, I. 1967. The ecology and etiology of human
disease. p. 106-107. Charles Thomas Pub. Co., Spring-
field, Ill.

49. Vaught, R., and A. Bleiweis. 1974. Antigens of
StAeptococcus mutans. II. Characterization of an antigen
resembling a glycerol teichoic acid in walls of strain
BHT. Inf. Immun. 9:60-67.




- 67 -


50. Vincent, J. M. 1970. A manual for the practical study
of root nodule bacteria. IBP Handbook No. 15. Blackwell
Scientific Pub., Oxford, England. 164p.

51. Weaver, R., and L. Frederick. 1972. A new technique
for most-probable-number counts of Rhizobium. Plant
Soil 36:219-222.

52. Wimalajeewa, D. L. S., and J. E. DeVay. 1971. The
occurrence and characterization of a common antigen
relationship between Ustitago maydis and Zea mays.
Physiol. Pl. Path. 1:523-535.

53. Worthington Biochemical Corporation. 1972. Worthington
enzyme manual. Worthington Biochemical Corporation,
Freehold, N. J. 216p.

54. Yao, P. Y., and J. M. Vincent. 1969. Host specificity
in the root hair "curling factor" of Rhizobium spp.
Aust. J. Biol. Sci. 22:413-423.

55. Zabriskie, J. B. 1967. Mimetic relationships between
group A streptococci and mammalian tissue. Adv.
Immunol. 7:147-188.













BIOGRAPHICAL SKETCH

Frank Bryan Dazzo was born April 8, 1948, in Miami,

Florida. He was graduated from Coral Gables High School in

Coral Gables, Florida, in June, 1966. In June, 1968, he

received an Associate of Science degree from Young Harris

College, Young Harris, Georgia, where he was elected to Who's

Who in American Junior Colleges. In August, 1970, he received

his Bachelor of Science degree with a major in bacteriology

from The Florida State University, Tallahassee, Florida. He

began his graduate studies at the University of Florida in

September, 1970, and received his Master of Science degree

in microbiology there in June, 1972.

In November, 1974, he was the recipient of the President's

Award from the Southeastern Branch of the American Society for

Microbiology for outstanding research presented at that meeting.

In December, 1974, he was awarded a Grant-in-Aid for Research

by Sigma Xi, the Scientific Research Society of North America.

-'T is a member of the American Society for Microbiology, the

Southeastern Branch of the American Society for Microbiology,

the Florida Soil and Crop Science Society, the Society of

Sigma Xi, and the National Honorary Society of Phi Kappa Phi.

He is currently a candidate for the Ph. D. degree in the

Department of Microbiology, University of Florida.

He is married to Olgalina G. Dazzo.


- 68 -








I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly pre-
sentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.




David h. Hudbell, Chairman
Associate Professor of Soil Microbiology


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly pre-
sentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.




Paul H. Smith
Professor of Microbiology


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly pre-
sentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.




Arnold S. Bleiweis
Associate Professor of Microbiology


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly pre-
sentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.




Edward M. Hoffmfann//
Associate Professo f Microbiology








I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly pre-
sentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.



tJames"t. Preston, III
Associate Professor of Microbiology


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly pre-
sentation and is fully adequate, in scope and quality, as a
dissertation for the degree of Doctor of Philosophy.



Raghavan Charudattan
Assistant Professor of Plant Pathology



This dissertation was submitted to the Graduate
Faculty of the College of Agriculture and to the Graduate
Council, and was accepted as partial fulfillment of the re-
quirements for the degree of Doctor of Philosophy.



December, 1975 O a r7

Deanollege of Agricu ture


Dean, Graduate School




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EODXUEMPI_E9D853 INGEST_TIME 2017-07-14T23:25:20Z PACKAGE UF00097524_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

CROSS REACTIVE ANTIGENS AND LECTIN AS DETERMINANTS OF HOST SPECIFICITY IN. THE RHIZ0BIUM-CLOVER SYMBIOSIS By FRANK BRYAN DAZZO A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1975

PAGE 2

UNIVERSITY OF FLORIDA 3 1262 08552 4261

PAGE 3

To my wife, Olga Exploration We shall never cease from exploration And the end of all our exploring Will be to arrive where we started And know the place for the first time Anonymous

PAGE 4

. ACKNOWLEDGEMENTS The author would like to express his deep appreciation to the chairman of his committee, Dr. David H. Hubbell for his constant guidance, concern, interest, and support. He would like to thank his committee members Drs. Raghavan Charudattan, Arnold S. Bleiweis, Edward M. Hoffmann, and James F. Preston for their assistance and suggestions and especially Drs. Paul H. Smith and Charles F. Eno for many years of friendship and support. He would like to thank Mr. Manuel Mesa, Mr. James Struble, and Dr. Willis Wheeler for their professional assistance and especially Ms. Carolyn Napoli, who has been a pleasure to work with throughout this study. He would also like to thank Ms. Napoli for providing the electron photomicrograph of Rk-izob^um tfiA.^olA.1 presented in Fig. 19. The author thanks the many scientists listed in Table 1 who provided the Rk-izoblam cultures used in this study. The author wishes particularly to express his loving gratitude to his wife, Olga, for her encouragement and understanding which made this dissertation possible. This research was supported by the National Science Foundation Grant No. GB-31307 and a Grantin-Aid for Research from Sigma Xi, the Scientific Research Society of North America.

PAGE 5

TABLE OF CONTENTS ACKNOWLEDGMENTS LIST OF TABLES LIST OF FIGURES ABSTRACT INTRODUCTION LITERATURE REVIEW MATERIALS AND METHODS . RESULTS AND DISCUSSION LITERATURE CITED BIOGRAPHICAL SXETCH . . Page , iii v vi .viii 1 2 5 . 18 . 63 . 68 IV

PAGE 6

LIST OF TABLES Table Page 1. Sources and Infectivity Characteristics of Rhizobiam Strains 6 2. Agglutination of Rki.zoblu.rn tli-iolli with Rabbit AntiTVu. ^ol-Lam n.e.pe.ni Root Antiserum ... 19 3. Reactivity of Anti-Tii&oliu.m fie.po.Yii> Root Antiserum and Rhi.zob Root Hairs 50 51 Inhibition of Clover Lectin-Mediated Cell Agglutination by Various Carbohydrates 53 Inhibition of Infection and Modulation of TnJL folium xzpz,\i> with Rklzoblam tfil^olli 403 by 2-Deoxygluccse 55 56

PAGE 7

LIST OF FIGURES Figure Page 1. Immunofluorescence of RhA.zob-iu.rn t>l root antiserum 32 9. High pressure liquid chromatography of the Rklzoblmm th.i{olii 403 capsular antigen on Bio-Glas 2500 34 10. Gas-liquid chromatography of trimethylsilylated sugars in an acid hydrolysate of Rhlzoblum tfiliolii 403 capsular antigen 36 11. Infrared spectra of Rhlzoblixm Vil^oHi 403 and Bart A capsular antigens 38 12. The effect of pH on the immunof luorescent cross reactivity of encapsulated cells of Rhizobium Vil^olll 403 with anti-T/t^o-tUum tepeni root antiserum 40 VI

PAGE 8

LIST OF FIGURES — Continued Figure Page 13. Tn.ifaoZium fan.agifaen.um root hairs .45 14. Tsi4.6ol.ium £ Jia.g4.fae.fi am root hair deformation in the presence of the capsular material from Rh4.zob.ium tn.i.faolii 403 (100 ug/ml) 45 15. Tn.ifaoli.um fan.ag4.faen.um root hair deformation in the presence of the capsular material from Rkizobium tuifaoLll Bart A (100 yg/ml) .... 45 16. Rhlzobium tn.ifaolii 2S-2 cells suspended in saline 48 17. Agglutination of Rkizobium Viifaolii 2S-2 cells by the Txifaolium nepeni seed extract 48 18. Adsorption of Rkizobium tnifaoiii 403 cells on a Tn.ifaol.ium n.epeyu> root hair 58 19. Electron photomicrograph of an ultrathin section of Rkizobium Viifaolii NA-3 in association with a Tn.ifaot4.um fan.agifazn.um root hair 58 20. Schematic diagram of the proposed cross-bridging of the cross reactive antigens of Rkizobium tn.ifaol.ii and Txifaotium root hair with a clover lectin 60 vii

PAGE 9

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CROSS REACTIVE ANTIGENS AND LECTIN AS DETERMINANTS OF HOST SPECIFICITY IN THE RHIZ08IUM-CLOVER SYMBIOSIS By Frank Bryan Dazzo December, 1975 Chairman: Dr. David H. Hubbell Major Department: Microbiology The basis for host specificity in the Rfu.zob.ui.mTn.A.faollu.m (clover) symbiosis was investigated. Cross reactive antigens of clover roots and R. .t-ii&ot-Li. were detected on their cell surfaces by tube agglutination, immunof luorescent and radioimmunoassay techniques. Anticlover root antiserum had a higher agglutinating titer with infective strains of R. tKl^ollL than with noninfective strains. The root antiserum previously adsorbed with noninfective R. tli-hol-li cells remained reactive only with infective cells, including infective revertants. When adsorbed with infective cells, the root antiserum did not react with either infective or noninfective cells. Other Rfuzob-cum species incapable of infecting clover did not demonstrate surface antigens cross reactive with clover. viii

PAGE 10

Radioimmunoassay indicated twice as much antigenic cross reactivity of clover roots and R. ti-L&ol-L-l 403 (infective) as compared with R. tH-l^aLLX. Bart A (noninfective) . Immunofluorescenceusing anti-R. tni. ^ o LLL (infective) antiserum was detected on the exposed surface of the root epidermal cells and diminished at the root meristem. The immuno fluorescent cross reaction on clover roots was totally removed by adsorption of anti-R. tfil^olll (infective) antiserum with encapsulated infective cells but not with noninfective cells. The cross reactive capsular antigens from R. tn.A.^ol-0. strains were extracted and purified. The ability of these antigens to induce clover root hair deformation was much greater when they were obtained from infective as compared with noninfective strains. The cross reactive capsular antigen of R. tn.i.^oli.1 403 was characterized as a high molecular weight (>4.6 x 10 daltons) , amorphous, B-linked, acidic heteropolysaccharide containing 2-deoxyglucose, galactose, glucose, and glucuronic acid. A soluble, nondialyzable substance (clover lectin) capable of binding to the cross reactive antigen and agglutinating only infective cells of R. th.A.^olil was extracted from white clover seeds. This lectin was sensitive to heat, pronase, and trypsin. Inhibition studies indicated that 2-deoxyglucose was the most probable haptenic determinant of the cross reactive capsular antigen capable of binding to the root antiserum and the clover lectin. Infection IX

PAGE 11

and nodulation of white clover roots by R. tKA.{oti.l was inhibited by 2-deoxyglucose at concentrations of at least 30 mM. The adsorption of infective and noninfective R. tfil^olll cells to white clover root hairs was examined. After 12 h of incubation, four to five times more infective cells than noninfective cells were firmly adsorbed to root hairs. A model is proposed to explain host specificity based on the preferential adsorption of infective versus noninfective cells of R. th.iholi.1 on the surface of clover roots by a cross bridging of their common surface antigens with a multivalent clover lectin.

PAGE 12

INTRODUCTION The N fixinq R'n-izob-iumlegume symbiosis is character2 ized by a high degree of host specificity controlled by the bacteria and the plant. Expression of host specificity is an early event which occurs prior to the formation of infection threads within the root hairs of the host (29) . The purpose of the study reported here was to determine the basis for host specificity in the Rk-Lzoblum tnliolliTfiliallmm (clover) symbiosis. It was hypothesized that R. tfiliolll cells have specific surface antigens which are involved in the adhesion of the bacteria to the cross reactive surface antigens of the clover root hair cell wall with the aid of a clover lectin. This association was chosen for study because of the small-seeded nature of the macro symbiont which was ideal for microscopic studies of infection and the availability of mutant strains of R. Vilbolll which have lost the ability to infect clover. 1 -

PAGE 13

LITERATURE REVIEW Cross Reactive Antigens The importance of common or cross reactive antigens in the microbial invasion of a host is receiving increasing recognition. Antigenic similarities between vertebrate hosts and microbes have been implicated in the pathogenesis of rheumatic fever (55) , human ulcerative colitis (41) , rat glomerulonephritis (34), and rat arthritis (6). In these cases, pathogenesis may involve immunological tolerance of the cross reactive antigens by the host or deposition of cross reactive cytotoxic antibodies on the host tissue. Common antigenic substances between invasive microbes and their plant hosts have also been found. These antigenically similar cell constituents possibly underlie host-pathogen compatibility based on their correlation with disease development. According to one theory (16, 17), a strong common antigen relationship between a plant host and a pathogen might result in the least disruption of cellular function between a pathogen and its host during infection with consequent success in disease development. Common antigen relationships have been implicated in the pathogenesis of flax rust by MzlampAOfLa llnl (18), angular leaf spot in cotton by Xanthoma nai malvac2.afiu.rn (17, 45), black rot of sweet potatoes by CVuUocy&tl& iimb.ila.ta. (17) , common smut of corn by Uttilago maijdli (52), crown gall tumor of 2 -

PAGE 14

3 tobacco by Kqn.oba.c.-tiLn.i.u.m tum2.frac-Le.nA ($) , and wilt of cotton by TuAcUiium and V2.MA.c-Lll-Lu.rn (9) . Host specificity is sharply defined for all of these phytopathogens except A. tumzfiaclzn* which has a wide host range. Cross reactive antigens have been found between 8 legumes and 3 species of RhJ.zob-tu.rn (10) . However, there was no correlation between the numbers of common antigens (immunoprecipitin bands in Ouchterlony plates) and the ability of the bacteria to infect their respective legume hosts. Lectins A lectin is a non-antibody protein or glycoprotein capable of specific interaction with carbohydrates. Lectins have been isolated from a variety of plants and non-vertebrate animals. Many lectins are phytohaemagglutinins since they can agglutinate erythrocytes by binding to their surface carbohydrate components. Interactions between legume lectins and Rh.-Lzob-iu.rn have been suggested (1, 48) as well as documented (4, 5, 7, 15, 22) . Hamblin and Kent (22) showed that phytohaemagglutinin l?ha.£>zolu.& lectin) could bind to R. phaAzol-L and that this lectin was present in the seeds, nodules, and on the roots below the nodules of ?h&&2.olu.t> vulgasi-L6 . Bohlool and Schmidt (4) demonstrated a high correlation between the binding capacity of soybean lectin to Rh-Lzob-Lum cells and the ability of these bacteria to nodulate soybean. They proposed that the legume lectin may serve as the basis for host specificity by interacting specifically with a polysaccharide on the surface of the Rh-Lzob-Lam cell. The

PAGE 15

4 lectin binding site on the RhA.zobA.um cell and the means whereby the lectin binds to the plant roots were not examined. However, some strains of Rhi.zobA.um do not bind to lectins obtained from the legume host that they nodulate (4 f 5) , and other strains of Rh.-Lzob4.um incapable of nodulating a certain legume still bind to the lectin from that host (5, 15). Thus, interactions between legume lectins and RhA.zobA.um cells may not always account for the specificity expressed by the nodule bacteria for their respective legume hosts (15) .

PAGE 16

MATERIALS AND METHODS Strains of RhJ.zob.lum The sources, infectivity, and legume hosts of the Rhlzoblum strains are listed in Table 1. Infectivity is defined as the formation of root hair infection threads in small-seeded legumes using glass slide assemblies (20) , or the production of root nodules on large-seeded legumes planted in cellophane pouches (51) . The spontaneous infective revertants of R. tn.lioli.1 (BA-L, BA-S, and 0435-21) were isolated from nodules of Tfil^olium Jizpzni inoculated with the corresponding noninfective strains. Preparation of Antigens Bacterial cells were grown on a modified Bergersen's chemically defined medium, harvested, and sonicated as previously described (14). Root antigens of T. -tepe/u var. Louisiana Nolin and T. ^icLglfrzium var. Salina (hereafter called white and strawberry clover, respectively) were nrepared. Seeds (50 g) were surface sterilized, spread on water agar plates, overlayed with sterile stainless steel wire mesh, inverted, and cold treated (38). The seeds germinated through the wire mesh into humid air at 22 C. After 3 days, seedling roots were excised along the wire mesh with razor blades and frozen in liquid N 2 Roots were macerated by grinding and thawing in 30 ml phosphate buffered saline (PBS, 0.05 M K 2 HP0 4 "KH 2 P0 4 , 0.15 M NaCl, 0.001 M MgS0 4 , 5 -

PAGE 17

6 4J

PAGE 18

7 >. •P •H > •H 4J O (1) M-l C H e n •H C 03 0> M O

PAGE 19

8 pH 7.2) containing .0.05 M sodium ascorbate (9) and 30 g of insoluble polyvinylpyrrolidone (PVP, Bio-Rad Laboratories, Richmond, Calif.) which had been previously washed with PBS 8 times to remove 28 nm absorbing material (13) . The slurry was stirred at C for 20 min, then centrifuged at 27,000 x g for 30 min. The supernatant was decanted, concentrated 6-fold by dialysis at 4 C against 20% (w/v) polyethylene glycol (PEG) dissolved in PBS (14), dispensed in 2 ml aliquots, and stored at -7 C for future use. Particulate antigens were also prepared by mascerating seedling roots in PBS containing 20% soluble PVP (40,000 av. mol. wt. , pharmaceutical grade, Sigma, St. Louis, Mo.) and then ultracentrifuging at 104,000 x g for 3 h. The pellet was washed twice with PBS, ultracentrifuged between washes, and then used as particulate antigens for immunization. Preparation of Antiserum RhA.zoblu.rn and clover root antigens were emulsified with equal volumes of Freund's incomplete adjuvant and used to prepare antiserum in virgin New Zealand white rabbits according to the immunization schedule previously described (14) S..eep anti-rabbit y globulin antiserum and fluorescein isothiocyanate (FITC) labeled y globulin fraction of goat anti-rabbit y globulin were purchased from Nutritional Biochemicals Co., Cleveland, Ohio, and Difco Laboratories, Detroit, Michigan, respectively. Serological Techniques Tube Agglutination Tube agglutination of Rki.zobi.um cells was performed

PAGE 20

9 using saline (0.15 M) as the diluent (50). Eight strains of R. tni^olil were examined, including 4 infective strains (2S-2, 403, T37, and 0435) and 4 noninfective mutants (2L, Bart A, Bio9, and 0435-2). Immunofluorescence Both the direct and the indirect immunof luorescent tests were run according to standard procedures (21) . Controls for detection of autof luorescence, nonspecific staining, and normal serum reactivity were included. Details of the optical system employed were described elsewhere (15) . Immunoprecipitation A quantitative precipitin test using the purified capsular antigen of R. trifolii strain 403 and rabbit anti-white clover root antiserum was performed according to Nowotny (40). The protein contents of the washed immunoprecipitates were determined with the Fclin phenol reagent (31) . The supernatants were analyzed for excess antigen or antibody using the capillary precipitin test. Radioimmunoassay For the radioimmunoassay, 10 yg (protein) of the -\ globulin fraction (21) of sheep anti-rabbit y globulin 125 were lodinated with 2 mCi of Na I (Amersham Searle, Arlington Heights, 111.) using the chloramine T method (25) . The iodinated protein was separated from free iodine by gel filtration on a 1 x 10 cm Sephadex G-50 column (Pharmacia, Uppsala, Sweden) . R. ifii^olil cells (strains 403 and Bart A) were suspended in PBS to a final density 9 of 1.6 x 10 cells per ml as determined by direct counting

PAGE 21

10 with a Petroff-Hausser chamber. Rabbit anti-white clover root antiserum (0.01 ml) was added to 0.05 ml of the cell suspensions run in duplicate, incubated at 37 C for 3 min, C for 3 min, and then brought up to 1 ml volume with PBS. The suspensions were washed twice with PBS by centrifugation at 1,000 x g for 10 min each. This procedure of adding cross reactive antibody, incubation and washing was repeated twice. The washed cells adsorbed with the cross reactive antibody were incubated with the 125 I labeled goat anti-rabbit y globulin added in 0.06 ml increments. The reaction mixtures were incubated for 30 min at 37 C, followed by 30 min at C. Following washing of cells as before, the pellets were counted for y emissions in a Model E116A-0 y counter (Nuclear Equipment Co., 125 Farmingdale, N. Y.) This procedure of adding the I labeled anti-y globulin was repeated until no further increase in antibody binding occurred, indicating antibody saturation. Tubes lacking anti-clover root antiserum were processed to quantitate non-specific binding of antiglobulin to cells, the inner wall of the tubes, and background -, radiation. Adsorption of Antiserum One ml of antiserum was incubated at C with 0.05 ml packed cell volume for 15 min with constant mixing, and then centrifuged at 1,500 x g for 15 min. The supernatant was readsorbed twice with fresh cells. The adsorbed antiserum was unreactive with the adsorbing cells as determined by indirect immunofluorescence. Encapsulated cells of R.

PAGE 22

11 tK-i^oL-LL strain 403 used for adsorption of cross reactive antiserum were separated from unencapsulated cells by differential centrifugation in PBS at 1,500 x g for 20 min. Clean, reliable separations were achieved under these conditions. Unencapsulated cells sedimented faster than encapsulated cells, forming a pellet under the diffuse layer of encapsulated cells. This upper layer was drawn off and centrifuged at 27,000 x g for 15 min to pellet the encapsulated cells. Analytical Studies of R/i-iz ob.iu.m Capsular Antigens Purification The capsular antigens of the R. th.i.{ i oti-i strains were purified by procedures using cetylpyridinium chloride (40) , and hexadecyltrimethyl ammonium bromide (11) as complexing agents. The material was lyophilized and stored in a desiccator for future use. Precipitation and Immunofluorescence The capsular antigen of R. tfil^alll 403 (hereafter called 403-AC) was tested for reactivity with 1% soluble diethylaminoethyl-dextran (DEAE-dextran) and 1% dextran sulfate (Pharmacia [19]), clover lectin (discussed later), and homologous and heterologous cross reactive antibody in capillary tubes. Some 403-AC was spread on microscope slides and tested for reactivity with rabbit anti-white clover root antiserum and anti-R. th.-i^ot-i.i 403 antiserum using the indirect immunof luorescent technique. Electrophoresis The capsular materials (2 mg/ml) from R. t-il^ol-ii

PAGE 23

12 403, and Bart A (hereafter called BAC), were spotted on cellulose polyacetate strips (Sepraphore III, Gelman Co., Ann Arbor, Mich.) and electrophoresed for 1 h at 22 C under constant voltage (300 V) at pH 8.6 (0.05 M sodium barbital) . The strips were dried, stained in 1% alcian blue in 0.01 N HC1 for 10 rain, and then washed in 0.01 N HC1 overnight. Gel Permeation Chromatography A sample of 403-AC was hydrated from the lyophilized state and was passed through a 0.20 um filter. An average molecular weight was determined by gel permeation (3) using a high pressure liquid chromatograph (Model 202/401 Waters Associates, Milford, Mass.) A sample of 403-AC was chroma tographed in two stainless steel columns (3.2 x 600 mm) connected in series and packed with porous glass beads (AX through FX, Waters Assoc, and Bio-Glas 2500, Bio-Rad Laboratories) deactivated with 1% PEG, and was detected by a refractive index monitor. Degassed water was the solvent. The columns were calibrated by determining the elution 3 6 volumes of soluble dextrans (5 x 10 to 2 x 10 daltons, a*, mol. wt. , Sigma) and purified plant viruses (Bromefi 7 grass mosaic, 4.6 x 10 , and tobacco mosaic, 3 x 10 daltons, courtesy of Dr. E. Hiebert, Dept. Plant Pathology, Univ. Fla.) Biochemical Component Analysis Samples of 403-AC and BAC were acid hydrolyzed (11) , converted to the volatile trimethylsilylated derivatives with Tri-Sil (Pierce Chem. Co., Rockford, 111. [43]), separated and detected by gas-liquid chromatography employing

PAGE 24

-inflame ionization detection as previously described (38). Sugar derivatives were identified by comparison with retention times of authentic standards (Nutritional Biochemicals Co..) The 4-0-methyl, D-glucuronic acid (24) was a gift from Dr. F. Loewus , Dept. Biology, State Univ. of New York (Buffalo, N. Y.); ct-D glucuronic acid was a gift from Calbiochem, La Jolla, Calif. The neutral sugar and uronic acid content of 4 03-AC was quantitated color imetrically by the anthrone and carbazole methods, repectively (28). Amino acid and amino sugar analyses of 4 03-AC hydroly sates were conducted on a JOEL Model JLC-6AH automated amino acid analyzer (JOEL, Inc., Cranford, N. J. [49]). Diffraction and Spectral Analyses X-ray diffraction and infrared spectral analyses of 403-AC and BAC were performed as previously described (38). A UV absorption spectral analysis of 403-AC dissolved in water (2 mg/ml) was performed using Suprasil cuvettes (Precision Cells Co., Hicksville, N. Y.) and a Beckman DBG grating spectrophotometer. Antigenic Cross Reactivity of R. VlUoUJ. 403 After Enzymatic Digestion Cells were fixed on microscope slides, then incubated 4 days at 22 C in the dark with the highest purity grade of pronase, trypsin, lysozyme, deoxyribonuclease, ribonuclease, pectinase, cellulase, and phospholipase dissolved in their appropriate buffers (Worthington Biochemical Corporation, Freehold, N. J. [53]). Cells were also treated with 7 M urea, 1% Triton X-100 (Sigma), 1% Tween-80 (Difco) , 0.5 M

PAGE 25

14 sodium periodate, aqueous HC1 (pH 3, 4, 5, and 6), deionized water (pH 7), and aqueous NaOH (pH 8, 9, 10, 11, and 12). Cells plus the enzyme buffers were also examined. After incubation, cells were washed with deionized water and tested for antigenic cross reactivity using anti-white clover root antiserum by the indirect immunof luorescent technique. Biological Assay The effect of the capsular material (100 ug/ml, [23]) on strawberry clover root hair development was determined as previously described (46) . Root hairs along the 4 optical median planes of two seedling roots per capsular preparation were examined by phase contrast microscopy. Hairs were counted, averaged, placed into straight, branched, moderately curled (<360°), or markedly curled (>360°) categories (54), and the data evaluated statistically. Clover Lectin Studies Extraction White clover seeds (50 g) were ground to a fine powder and mixed with 150 ml PBS plus sodium ascorbate (0.05 M) with further grinding. Washed, insoluble PVP (20 g) was added, the slurry was stirred at 4 C for 1 h, filtered through PBSwashed cheese cloth, and centrifuged at 27,000 x g for 1 h. The upper lipid layer was removed, and the remaining supernatant was passed through 0.20 ym membrane filters (TCM-200, Gelman) . This seed extract (hereafter called LAI) contained 10 mg/ml protein after dialysis against PBS to remove ascorbate. LAI was stored in 0.5 ml aliquots at -20 C for future use.

PAGE 26

15 Binding to Rh.ZzobJ.um Cells Cells washed in PBS were tested for evidence of binding with LAI by immunofluorescence, tube agglutination, and slide agglutination. Since several Rk^zobium strains flocculate spontaneously due to cellulose microfibril production (38), uniform cell suspensions were obtained by filtering cells in PBS through glass wool loosely packed in Pasteur pipettes (5 x 10 mm) . Cell suspensions were added to equal volumes of LAI, incubated at 30 C (tubes) or 22 C (slides) for 4 h, and then examined for agglutination. In other studies, cells were mixed with LAI, incubated for 4 h, and washed twice with PBS in the centrifuge at 1,400 x g. Antigenic material in LAI capable of firmly binding to cells (not removed by washing) was detected by agglutination and indirect immunofluorescence using rabbit anti-LAI. This antiserum was prepared according to the immunization schedule for clover root antiserum and by itself was unreactive with Rhizoblum cells using tube agglutination and immunofluorescence. Characterization of Agglutination Factor Dialysis . The agglutinating titer of LAI using R. th-l^oHl strains 2S-2 and 403 was determined. An aliquot (3 ml) of undiluted LAI was dialyzed against PBS (3 changes in a total of 15 liters) and then titrated. Ultracentrifugation . The agglutinating titer of LAI before and after ultracentrifugation at 104,000 x g for 1 h was determined.

PAGE 27

16 Heat lability . ' Aliquots of LAI were heated at 56 C or 8 C for 10 min, filtered, and titrated. Enzymatic digestion . Aliquots of LAI were diluted 1:8 (endpoint agglutination titer) by adding the various enzymes and other reactive materials in PBS as described above. These digestion mixtures were incubated for 24 h at 3 C, centrifuged at 3,000 x g for 10 min, and the supernatants were tested for agglutination of R.t
PAGE 28

17 inoculated with 2.6 x 10 R. VuLioLU. strain 403 cells were prepared with various increments of sterile 2-deoxyglucose (final concentrations of 0, 2.5, 5, 10, 30, 50, 100, and 200 mM) or a-D glucose (200 mM) in duplicate. Slides were incubated for 20 days in a plant growth chamber (Warren Sherer, Model CEL 255-6, Marshall, Mich.) programmed at 22 C isothermal, 12 h photoperiod, 18.6 lux light intensity, and then examined for evidence of root hair infection by phase contrast microscopy. Adsorption of R. til^olZl Cells to Root Hairs of T. n. e. p e n-i, Slide assemblies of T. >i2.pzn& seedlings without agar g were inoculated with approximately 1.9 x 10 infective or noninfective R. tn.X.^otX.Ji cells. After 12 h incubation in the growth chamber (10 h darkness followed by 2 h light) , the slides were removed. The cover slips were removed, then the roots were washed on the slides by a gentle stream of saline. The slide assemblies were filled with saline, new cover slips were added, and then the roots were examined by phase contrast microscopy. Four optical median planes of two primary roots per strain were evaluated. The number of bacterial cells per strain adsorbed to root hairs approximately 200 urn in length was determined and the data were evaluated statistically. Electron microscopic examination of R. tfiA.^oZ-LA. associated with T. j$A.agi.Q eAum root hairs was performed as previously described (38) .

PAGE 29

RESULTS AND DISCUSSION Antigenic Cross Reactivity between R. tfiljoUl and Clover Roots Tube Agglutination Antiserum to white clover roots cross agglutinated both infective and noninfective R. Vil&olli strains (Table 2). Endpoint cross agglutination titers were higher for the infective strains in 3 out of 4 infective-noninf ective strain combinations. In the fourth combination, both the infective and noninfective strains had equal agglutination titers. Cells were not agglutinated in normal serum controls. These results indicated that cross reactive antigens between clover roots and the surface of R. tllioll-i strains exist, and that the surface antigenic composition of the infective strains may be different from the noninfective strains. Immunofluorescence using Anti Clover Root Antiserum Unadsorbed anti-white clover root antiserum was reactive with the surfaces of infective and noninfective R. tll^oLX.1 cells when examined by indirect immunf luorescence (Table 3, Fig. 1) . The anti-white clover root antiserum was adsorbed with cells of each combination of infective and noninfective strains of R. tilholli. When adsorbed with whole cells of the noninfective strains, this antiserum remained reactive only with infective ceils, including the three spontaneous infective revertants 0435-21, BA-L and BA-S (Table 3). These three revertants also reacted strongly with antiserum against

PAGE 30

19 Table 2. AggLnti.na.tton ofc Rh-izobium ttLJL&oJLJLA. wtth Rabbet AnttT>i<. ft otlvim n.zpe.n-i> Root Ant-Li e.A.um Strain Endpoint titer 128 128 128 8 64 4 64 8 a 2S-2 a

PAGE 31

20 Table 3. Rzactiv-ity o& KntL1 'Ki.^0 IZum Jizpzni Root Antiizium and Rn. 'ND, not determined. Grown on a chemically defined medium. 'Grown in soil extract.

PAGE 32

Fig. 1. Immunofluorescence of Rkizobium Villolll 403 (infective) using anti-T/uc^o^um lapzn* root antiserum. Bar scale equals 2 um.

PAGE 33

-22-

PAGE 34

23 the parent infective strains. When adsorbed with cells of the infective strains, the root antiserum was reactive with no no. of the R, th.i.{oLJLX. strains examined. These results indicated that R." tfii^otll cells have antigens on their surfaces which are cross reactive with clover roots, and there is a distinctly greater degree of antigenic cross reactivity displayed by the infective strains. A portion of this "homology" is lost when cells lose the ability to infect the root hairs of their host, but it is reacquired when the cells spontaneously revert back to their infective state. R. tn.JL{olll strains Jl and J2, which were recently isolated from root nodules of natural stands of white clover, were also reactive with the anti-white clover root antiserum. This indicated that the antigenic cross reactivity of R. th.X.ioli.A. and clover is not unique only to laboratory strains maintained for years as stock cultures. While growing in soil extract, R. t>il{oLX.i. NA-30 (infective) was capable of maintaining a cell surface which was antigenically cross reactive with white clover roots. Controls for autof luorescence of cells and non-specific staining of FITC-labeled anti-rabbit y globulin were negative (no fluorescence) . Preimmune sera were unreactive. Anti-white clover root antiserum was tested for reactivity with a variety of other RhZzobium species, including R. japonicum, R. Izg umi.no 4 antim , R. phaszoli, R. me.ZiJ.otA., and members of the so-called "cowpea miscellany, "none of which infected clover (Table 1) . None of these RhizobZu-m species were reactive with antiserum to T. fio.pzn& (immunofluorescence) except R. sp. strain

PAGE 35

24 HR1. In this latter case, the preimmune normal serum control was equally reactive, and therefore the rabbit had natural antibody reactive with surface antigens of this organism. These results indicate that strains of Rhlzoblixm incapable of infecting clover lack specific surface components which are antigenically cross reactive with clover roots. Antisera were prepared against 27,000 x g supernatant and 104,000 x g pellet fractions of macerated root antigens of strawberry clover. Antisera to both fractions reacted with infective strains of R. ViJLioLU. (2S-2, 403, T37, and 0435) . This indicated that R. tn.X.{ol-il was antigenically cross reactive with at least 2 host species in the clover cross-inoculation group. It seems feasible to predict similar results with other clovers that can establish successful symbiotic relationships with R. tri-l^oHi. Cross reactive strawberry clover root antigen (s) remained in the supernatant at 27,000 x g for 1 h, and were sedimented at 104,000 x g for 3 h in sufficient quantity to elicit an immune response. A higher degree of antigenic similarity exists among legumes in the clover cross-inoculation groups (1) . K dip immunoassay Radioimmunoassay was performed to detect and quantitate the degree of antigenic cross reactivity between R. tn.i.^oHX.1 strains and the roots of their clover host. The amount of cross reactive antibody bound to equal numbers of strain 403 (infective) and Bart A (noninf ective) cells after repeated additions is shown in Fig. 2. At saturation, approximately twice as much cross reactive antibody bound to the infective

PAGE 36

25 Antibody Addition (0.06ml increments) Fig. 2. Radioimmunoassay of anti-TA-i folium nzpzyii, root antiserum bound to Rhi.zobium tul^olll 403 (infective) and Bart A (noninf ective) cells. Cells of both strains were adjusted to equal population sizes following direct microscopic counting using a Petrof f-Hausser chamber. Cells were incubated with rabbit anti-T. tie.pe.n.4 root antiserum, washed, and then incubated with L I-labeled goat anti-rabbit y globulin. After each addition, cells were incubated, washed, and counted for y emissions. Points and bars represent means and standard deviation.

PAGE 37

26 as compared with the noninfective cells. These data confirmed the earlier observations (tube agglutination and immunofluorescence) that quantitative differences in the degree of antigenic cross reactivity exist between clover roots and infective strains of R. tuX.^otUi as compared with noninfective strains. Immunofluorescence using Anti -R. tn.l^olli Antiserum Axenically grown white and strawberry clover seedling roots were reacted with antisera prepared against several infective R. tnlbolli strains (2S-2, 403, T37, and 0435), and then examined for the presence of bound cross reactive antibody using the indirect immunof luorescent technique. Antibody to each of the infective strains bound to the surface of sterile clover roots. Immunofluorescence by anti-0435 antibody appeared brightest at the growing root hair tip of white clover as compared with other external root parts. In all cases, immunofluorescence was present on the exposed surface of the root epidermal cells including root hairs (Fig. 3) and diminished at the root meristem. These results indicated that clover has antigens on its e. .posed root surface which are cross reactive with antigens from infective R. tn.JL^oti.X. cells. The blue autof luorescence of fresh root tissue (27) was removed by a K490 barrier filter. Jones and Russell (27) used immunofluorescence to identify R. tfi-i^olii. strains in T. n.zpzn.6 nodules. Interestingly, their photographs provided clear evidence of antibody to R. tn.-lkoLi.JL binding to surface antigens of clover roots without the authors actually reporting the observation.

PAGE 38

Fig. 3. Immunofluorescence of a T^^oHum izpcn* root using anti-RUz oblum tilbolll 2S-2 antiserum. Bar scale equals 45 ym.

PAGE 39

-28-

PAGE 40

29 Antisera to the infective strains of R. ViJi^oti.Ji previously adsorbed with whole cells of the corresponding noninfective strains bound to the antigens of the white clover root surface and the homologous infective R. tii&ol-t-L strains. Antibody to R. tn.i.^oLA.A. 403 (infective) which was cross reactive with the surface of clover roots could be removed by adsorption with encapsulated 4 03 cells. These results indicated important differences in the degree of antigenic cross reactivity between infective vs. noninfective R. tn.i.^0 l-ii strains and the roots of a clover host. Analytical Studies of R/i-iz ob-ium Capsular Antigens Since the cross reactive antigen of R. tK.JL^oLX.1 403 was localized on a morphologically distinct capsule as viewed by immunofluorescence, the capsular material was isolated and characterized. Precipitation and Immunofluorescence When dissolved in water, 403-AC was optically clear (Fig. 4). The 403-AC and the capsular material from Bart A (BAC) complexed with soluble DEAE-dextran (Fig. 5) , cetylpyridinium chloride, hexadecyltrimethyl ammonium bromide, and axcian blue. These compounds form specific precipitates with acidic polysaccharides (11, 19, 35, 40). 403-AC and BAC did not react with dextran sulfate which precipitates basic polysaccharides (19) . The 403-AC also formed a precipitate when reacted with anti-white clover root antiserum (Fig. 6) and a clover seed extract (Fig. 7, discussed later). The quantitative precipitin curve (Fig. 8) indicated that 403-AC was cross reactive with white clever roots. A maximum of

PAGE 41

Fig. 4. Rhlzoblum ttiliolll 403 cap'sular antigen dissolved in water (2 mg/ml) . y Fig 5. Reaction of Rhlzoblam tnlkolll 403 capsular antigen with DEAE-dextran. psuiar a „^ Fi ?: 6 * , R ^ action of Khlzoblmm tM{olli 403 capsular antigen with anti-T folium Ke.pzn& root antiserum. Fig ; Z' . Reaction of Rhlzoblum tMltolll 403 capsular antigen with T^^oUum *epen4 seed extract. capsular

PAGE 42

-31-

PAGE 43

32 Antigen Adiad Crag) Fig. 3. Quantitative precipitin curve of Rhlzoblum Vilbolli 403 capsular antigen with anti-Til&olium HQ.pQ.n6 root antiserum. Symbols at the top represent excess antxgen or antibody in the supernatants after removal of the lmmunoprecipitates by centrifugation.

PAGE 44

33 930 yg cross reactive anti-white clover root antibody protein was present per ml of undiluted antiserum. Antibody against white clover roots and R. tni^olil 403 reacted with 403-AC using immunofluorescence. Electrophoresis At pH 8.6, 4 03-AC and BAC were electrophoretically homogeneous. They both had a net negative charge, and migrated as single bands of equal electrophoretic mobilities (2.08 x 10~ 5 cm 2 /V sec). Gel Permeation Chromatography The capsular antigen 403-AC eluted as> a sharp peak just after the void volume of the glass bead columns of AX through FX (Waters Assoc.) and Bio-Glas 2500 (Bio-Rad, Fig. 9) No additional peaks of lower molecular weight in 4 03-AC were eluted within the selective permeation range of the beads. When compared with the elution volumes of various markers, these results indicated that 4 03-AC had an average molecular weight in excess of 4.6 x 10 6 daltons. This material was not contaminated with detectable amounts of any smaller compounds that would have changed the refractive index of water, b iochemical Component Analysis The capsular antigens 4 03-AC and BAC consisted primarily of carbohydrate. Total neutral sugar (anthrone) and uronic acid (carbazole) contents of 403-AC were 68.0 + 0.5% and 32.0 + 0.5%, respectively. Gas-liquid chromatography of the trimethylsilylated (TMS) derivatives of the sugars released by acid hydrolysis of 403-AC indicated 2-deoxyglucose,

PAGE 45

34 x u Q Z o < U. Bio-Glas 2500 HPLC-Gel Permeation of 403AC Vo TMV BMV DEXTRAN Fig. 9. High pressure liquid chromatography of the RkA.zobi.iLm tn-ifiol-Li. 403 capsular antigen on 3io-Glas 2500. Tobacco mosaic virus (TMV) , Bromegrass mosaic virus (BMV) , and dextran markers are 3 x 10 7 , 4.6 x 10 6 , and 2 x 10^ daltons, respectively.

PAGE 46

35 a-D galactose, a and (3-D glucose, a-D glucuronic acid, and one unidentified compound (Fig. 10). The unidentified compound (peak 2) did not correspond to the trimethylsilylated derivative of L-sorbose, L-fucose, 2-deoxyribose, 2-deoxygalactose, 3-O-methyl D-glucose, D-ribose, 6-deoxy-l-galactose, D-mannose, D-fructose, L-arabinose, D-xylose, 4-O-methyl D-glucuronic acid, D-melibiose, or L-rhamnose. Hexamethyldisiloxane was also present in the mixture as a normal by-product of the silylation reaction (43) . BAC gave an identical chromatogram (not shown) . The absence of deoxyribose and ribose indicated a lack of nucleic acid contamination. The analytical system employed could detect 1 nanogram of a-D glucose TMS as a lower limit of sensitivity. Trace amounts (1-10 nanomoles per mg dry wt) of lysine, aspartic acid, threonine, serine, glutamic acid, glycine, alanine, and isoleucine were identified in acid hydrolysates of 4 03-AC by comparison with — ratios and 570nm identical retention times of authentic standards. The total amino acid content could account for approximately 0.26% of the dry weight of 403-AC and therefore was not considered a major structural component of the capsule. Neither glucosamine nor galactosamine was detected within the sensitivity of the amino acid analyzer employed (10 nanomoles) . Diffraction and Spectral Analyses X-ray dif f ractograms of 403-AC and BAC revealed no reinforcement peaks within the range 2° 29 to 60° 20. This

PAGE 47

36 OLC of 403 AC HYDROIYSATE TMS RETENTION Fig. 10. Gas-liquid chromatography of trinethylsilvlated sugars in an acid hydrolysate of RkA.zobA.um tfii^olA-A. 403 capsular antigen. Peak 1= pyridine solvent, peak 2= unidentified, peak 3= 2-deoxyglucose, peak 4= a-D galactose, peak 5= a-D glucose, peak 6= hexadisiloxane, peak 7= 3-D glucose, peak 8= a-D glucuronic acid.

PAGE 48

37 indicated an amorphous, non-crystalline structure, and the lack of detectable contamination with cellulose microfibrils which were produced by both strains (38). The infrared spectra of both materials were essentially identical (Fig. 11) . The spectra were consistent with a carbohydrate structure rich in hydroxyl and carboxyl groups. The absorption band at 890-900 cm -1 was characteristic of a 8 glucosidic linkage and absence of a band at 870 cm -1 indicated absence of an a glucosidic linkage (2). The UV spectrum of 403-AC indicated absorption in the short UV range only (210-230 nm) . Lack of absorption down to 230 nm indicated no detectable nucleic acid and/or protein which would absorb in this range. It is concluded from these analytical studies that 403AC is a water soluble, amorphous, high molecular weight, 8-linked, acidic heteropolysaccharide. Antigenic Cross Reactivity of R. tfiljolll Strain 4 03 After Enzymatic Digestion Encapsulated R. tfil^olll 403 cells were treated with various enzymes and other reactive materials and then tested for cross reactivity with anti-white clover root antiserum using immunofluorescence. The antigenic cross reactivity of strain 403 capsules was eliminated by lysozyme, periodate, acid (HC1, pH 3) and alkaline (NaOH, pH 12) treatments. Lysozyme treatment lysed the cells. Apparently the capsular antigen was washed away from the cell walls solubilized by lysozyme. The capsules of the cells were removed by washing after periodate treatment. However, the underlying cell

PAGE 49

38 INFRARED SPECTRA 35OO WOO MOO MC WAVENUMBER (cm"') Fig. 11. Infrared spectra of RkZzobZum th.A.§oL-li. 403 and Bart A capsular antigens.

PAGE 50

39 surface remained antigenically cross reactive with clover roots. This indicated that these cells maintained an antigenically cross reactive surface even if they lost their capsules. Maintaining an underlying cross reactive layer could possibly be important for the cell in the rhizosphere if the C/N ratio of root exudates restricted abundant capsule formation. Effect of pH on Antigenic Cross Reactivity The antigenic cross reactivity of R. tfii^olil 403 cells following acid and alkaline treatment was examined in more detail. The degree of antigenic cross reactivity remained high in the pH range 6-8, but diminished outside of this range (Fig. 12). The acid sensitivity of the cross reactive surface antigen (at pH 5 and below) corresponded to the acid-sensitive step where root hair curling, infection, and nodulation abort (36). If possession of the cross reactive antigen is essential to the infection process as proposed here, then its loss at high H + concentration may provide a biochemical explanation of nodulation failure in acid soils despite proper inoculation. £.-• ological Activity of R. tKljolll Capsular Antigens Rhlzobium extracellular nondialyzable material has been known to induce deformation of legume root hairs (23, 30, 33, 54). A markedly curled deformation of roct hairs has been reported to be restricted to inoculation of a legume with live, virulent, homologous Rhizobium cells (54). Other have shown that root hair deforming substances produced by

PAGE 51

40 Fia 12 The effect of pH on the immuno fluorescent cross reactivity of encapsulated cells of ^hlzoblam VfifoU*. 403 with antL-TJilioUum MpinA root antiserum The K470 barrier filter was used to differentiate reactive and unreactive cells.

PAGE 52

41 R. th-t^oL-li. could be adsorbed to white clover roots and eluted again by acetic acid or urea (46) . The antigenically cross reactive material, which was isolated by the method of Nowotny (40) from infective and noninfective cells grown on a chemically defined medium, was tested for its ability to deform root hairs of strawberry clover grown axenically. The results are reported in 2 Table 4. Ax test indicated a highly significant dependence (at 99.9% level) of root hair deformation on the addition 2 of this capsular material. Another x test indicated (at 99% level) that the capsular material from' all the strains examined did not induce root hair deformation to the same 2 degree. Several x tests were performed to determine if the degree of deformation was independent of whether or not the curling factor came from the infective or the corresponding noninfective strain. These tests were all rejected at the 99% level, indicating that the degree of root hair deformation induced by the curling factor was significantly dependent on the strain from which it was isolated. Onetailed Z tests were performed to test if the frequency of observations within a deformation category (straight, branched, moderately or markedly curled) was greater for one strain than for another (e.g. , infective vs. noninfective) . The frequency of straight root hairs was significantly less (at 99% level) when incubated with the capsular material of the infective strains as compared with the corresponding noninfective strains. In all cases except the branching

PAGE 53

42 •^ o>3.-t? £

PAGE 54

43 frequency of 2S-2 and 2L, the degree of root hair deformation was significantly greater for the -In^zctZv e strains. Thus, the antigenically cross reactive capsular material of these R. tiifao ZZ'-l strains possessed biological activity characterized by the ability to deform root hairs of the clover host. Photomicrographs of control root hairs and root hairs deformed by 4 03-AC and BAC are presented in Figs. 13, 14, and 15, respectively. Clover Lectin Studies A clover lectin capable of differentiating infective and noninfective R. tfii^oll-L strains was sought. A heat sensitive, nondialyzable , soluble protein or glycoprotein agglutinating factor with definable sugar specificity has been found in LAI. The agglutinating factor reacted with the cross reactive surface antigens which these symbionts share. Binding to Rh-Lzob-lum Cells The spectrum of cell agglutination by LAI (Table 5) included all R. tsi-i&ol-Li strains capable of infecting white clover (including the infective revertants) but none, of the noninfective R. th-i^ol-ii. strains or other Rin.yizoblu.rn species incapable of infecting this legume. R. le.gumZno6a.Aum 3H0Q1 and R. sp. HRl could not be examined due to autoagglutination. Control and LAI-agglutinated cell suspensions of R. tni^olU 2S-2 are illustrated in Figs. 16 and 17, respectively. The ability to agglutinate indicated that the proposed clover lectin was at least divalent, i.e., had at least two reactive binding sites.

PAGE 55

Fig. 13. TtiZ&oli.um ^ftagZ^zftam root hairs. Bar scale equals 45 ym. Fig. 14. TfiZ^ol-ium ^ag^L^zKum root hair deformation in the presence of the capsular material from Rk-izoblam tnl^olll 403 (100 yg/ml) . Bar scale equals 45 ym. Fig. 15. Tfi-i^olZum ^n.ag-1^ znu.m root hair deformation in the presence of the capsular material from Rk-izobZum tfil^oll-i Bart A (100 yg/ml) . Bar scale equals 45 ym.

PAGE 56

-45• •>

PAGE 57

46 c o •H *J rd C •H -P rH CP en en 0) -H o a) ft CO e> c> O X! XI X! V£3

PAGE 58

Fig. 16. Rhlzoblum tfil^ol-il 2S-2 cells suspended in saline. Bar scale equals 0.5 mm. Fig. 17. Agglutination of Rkizobium ttii^oLLl 2S-2 cells by the T ' n.l^oll\xm A.epe.n-4 seed extract. Bar scale equals . 5 mm.

PAGE 59

-48-

PAGE 60

49 Whether the agglutinating factor could bind to cells at levels below the sensitivity of the agglutination test was then tested. Cells were incubated with LAI, washed, and then reacted with anti-LAI using tube agglutination and indirect immunofluorescence. Antigenic material in LAI capable of firmly binding to the cells was detected on both infective and noninfective R. tfil^olll strains but not on other Rhlzobiixm species incapable of infecting white clover (Table 6) . Indirect immunofluorescence showed that the proposed clover lectin bound to the distinct capsules of R. th-ilolli strain 403. Characterization of the Agglutination Factor The endpoint agglutinating titer of LAI did not change following extensive dialysis (3,500 molecular weight cutoff) and ultracentrifugation at 104,000 x g for 1 h. Treatment of LAI at 56 C and 80 C for 10 min completely destroyed its ability to agglutinate infective strains of R. tfil^olll. These results indicated that the agglutinating factor was nondialyzable, nonparticulate (soluble) , and heat labile. The effect of various enzymes and other reactive materials on the agglutination of R. ViJi^otll 403 cells by LAI was examined. Pronase, trypsin, periodate, acid (HC1, pH 3), alkaline (NaOH, pH 12), and 7 M urea treatments of LAI destroyed its ability to agglutinate the cells (Table 7). This indicated that the agglutination factor (lectin) in LAI was an acid and alkali sensitive protein (or

PAGE 61

50 -

PAGE 62

51 Table 7. E^zct o (J Va.Jii.oui> Tn&atm
PAGE 63

52 possibly glycoprotein) which may require the participation of H-bonding for agglutinability . These experiments did not distinguish between the possible need of H-bonding for intramolecular stability of the lectin and for intermolecular binding of the lectin to the surface of the RU.zob.tum cells. Periodate, acid, and alkaline treatments rendered the 4 03 cells non-agglutinable by LAI as with the anti-white clover root antiserum as described earlier. Clover root exudate and soil extract did not prevent the agglutination of R. tui-iolll 403 cells by the clover lectin. R. tKliolll NA-30 was also agglutinated by clover lectin after growth for one generation in soil extract. Sugar Binding Specificity Inhibition of LAI-mediated agglutination of R. tn.l^oli.1 403 and 2S-2 cells by various carbohydrates was examined. The sugar inhibition patterns were identical for both infective strains. Preincubation of LAI with 30 mM 2-deoxyglucose or N-acetylgalactosamine inhibited its ability to agglutinate the cells (Table 8). None of the other carbohydrates at this concentration were inhibitory. LAI agglutinated rabbit erythrocytes coated with 403-AC only if followed by anti-LAI. This antiserum did not agglutinate cells coated only with 403-AC. LAI-mediated passive haemagglutination was inhibited by 3 mM 2-deoxyglucose and N-acetylgalactosamine (Table 8) . Identical results were obtained with anti-white clover root antiserum. Since 2-deoxyglucose was a component of this cross reactive

PAGE 64

53 -

PAGE 65

-54capsular antigen, the sugar binding site of both the clover lectin and the anti-white clover root antibody was probably directed toward this sugar. Inhibition of Infection of T. Ktpzn* by 2-Deoxyglucose Since 2-deoxyglucose inhibited the binding of the clover lectin to the cross reactive antigen of R. tfii.^0 l-i-l , it was hypothesized that this sugar may interfere with the bacterial infection of clover by occupying all the lectin binding sites. Infection of white clover root hairs by R. tfil^ol-i-i 403 was completely inhibited by 2-deoxyglucose at concentrations of 30 mM and above (Table 9) . Partial inhibition was evident at lower concentrations (2.5-10 mM) . The inhibition was specific for 2-deoxyglucose and was not due to hypertonic osmotic effects since infection and nodulation occurred in the presence of 200 mM a-D glucose. However, 2-deoxyglucose is known to interfere with glucan wall synthesis, microfibril production, and cell metabolism in Saccha./iomyc^a.
PAGE 66

55 Table 9. lnh-i.bi.ti. on o^ Tn&e.cti.on and Nodu.la.tion o & Tn.i.{)OZi.am sicpun* with RhJLzobium

PAGE 67

56 Table 10. Kdlonptlon o i Rhizoblum Vtl^olli Czll*>

PAGE 68

57 the square root transformation (47) of the means. For every strain combination examined, the mean number of infective cells adsorbed was significantly greater (at 99.5% level) than the noninfective cells. Of particular importance was the finding that the infective revertant BA-L had a significantly greater frequency of adsorbed cells than the corresponding noninfective strain Bart A. Thus, it can be concluded that the phenotypic trait of infectivity for R. tn.i.^oLX.i. is directly correlated with the extent of cell adsorption to root hairs of the clover host. Another study has indicated no correlation between the extent of root adsorption with the infective capabilities of the microorganisms (42) . This study was based on indirect measurements involving viable colony counts of cells following their removal from roots by shaking. Unfortunately this latter approach did not distinguish root hairs from other adsorptive root surfaces, nondispersable floes from single cells, and the variability of root surface areas among individual plants. Adsorption of the rod-shaped bacteria occurred primarily through attachment of one of their poles to the root hair surfaces (Fig. 18) . Polar orientation has been reported as the position of attachment of single cells of R. tKllol-Ll (38), R. md.lX.Lotl (12), R. japon-lcum (44), and R. sp. (Ae.6ckynomzne. nodulating strain, [37]) on their appropriate legume host root surface. Ultrastructure of Adsorption The ultrastructural details of the association of

PAGE 69

Fig. 18. Adsorption of Rh.izobA.um tul^olll 403 cells on a JKl^ollum frnpzyii, root hair. Cells are polar ly attached, Bar scale equals 20 ym. Fig. 19. Electron photomicrograph of an ultrathin section of Rhizcbium tKl^olll NA-30 in association with a Jul ho Hum ifidQl^zfium root hair. The fibrillar capsule (A) of the bacterium is in contact with electron dense particles (B) on the outer periphery of the root hair cell wall (C) . Bar scale equals 1 ym.

PAGE 70

-59'-• 19

PAGE 71

60 R. tiibolii. NA-30 with the clover root hair wall are shown in Fig. 19. This photograph was typical of the polar orientation that was routinely observed. The fibrillar capsule of the bacterium was in physical contact with electron dense, globular aggregates lying on the exterior periphery of the fibrillar root hair cell wall. Based on the data of these investigations, the following model was proposed (Fig. 20) to explain this preinfective adsorption event which contributes to host specificity in the Rk-izo 6-tumclover association. Infective strains of R. tfii.boZ.-li have on their surfaces a polysaccharide which is antigenically cross reactive with an antigen on the root surface of the clover host, and also is capable of deforming root hairs. It is proposed that the clover lectin which recognizes these surface antigens cross-bridges them to form a correct molecular interfacial structure which allows for specific adhesion of the bacteria to the root surface. Following this specific cell adhesion, the invagination process of the root hair wall begins, resulting in infection thread formation (39). Noninfective R. tii.bolii. cells have cross reactive surface antigens but either in reduced quantity or sterically blocked as neither antibody nor lectin binds to them as efficiently as to infective cells. Other Rhlzobium species lack this surface cross reactive antigen, and therefore do not bind to the clover lectin. The model is consistent with the findings of Hamblin and Kent (22), and Bohlool and Schmidt (4) who introduced the role of lectin in the adsorption of the RiiA.zobi.am. But the

PAGE 72

61 Root hair cytcplcsm Fig. 20. Schematic diagram of the proposed crossbridging of the cross reactive antigens of Rk-czob-cum tn.^^oLi.1. and Ttii-Qolium root hair with a clover lectin.

PAGE 73

62 lectin alone does not confer specificity. It is one component of a multimembered specificity-determining complex. Two other components of this complex are the cross reactive surface antigens:

PAGE 74

LITERATURE CITED 1. Baldwin, I. L. , E. B. Fred, and E. G. Hastings. 1927. Grouping of legumes according to the biological reactions of their seed proteins. Bot. Gaz. 83:217-243. 2. Barker, S., E. Bourne, and D. Whiffen. 1956. Use of infrared analysis in the determination of carbohydrate structure. Meth. Biochem. Anal. 3:213-245. 3. Belue, G. P., and G. D. McGinnis. 1974. High pressure liquid chromatography of carbohydrates. J. Chromatogr. 97:25-31. 4 Bohlool, B. B., and E. L. Schmidt. 1974. Lectins: a possible basis for specificity in the Rklzob-tum-legume root nodule symbiosis. Science 185:269-271. 5. Brethauer, T. S., and J. D. Paxton. 1975. Interaction of soybean lectin with extracellular material from Rkizobium bacteria. Abstr. North Central Division Meeting, Amer. Phytopathol. Soc . , June 23-24, 197 j, West Lafayette, Ind. 6. Cahill, J. F., B. C. Cole, B. B. Wiley, and J. R. Ward. 1971. Role of biological mimicry in the pathogenesis of rat arthritis induced by Mycoplasma ah.tkn.itidA.& . Inf. Immun. 3:24-35. 7 Cappelletti, C. 1926. The bacteroid-like form and immunity in leguminous plants. Internatl. Congr. Plant Sci., Ithaca, N. Y. Proc . 1:59-60. 8 Chada, K. C, and B. I. Strivastava. 1971. Evidence for the presence of bacteria-specific proteins in sterile crown gall tumor tissue. Plant Physiol. 48: 125-129. 9. Charudattan, R. , and J. E. DeVay. 1972 Common antigens among varieties of Go66tjplum klA.Su.tam and isolation or Tu.ta.xium and V QjitlcilUam species. Phytopathology 62:230-234. in charudattan R , and D. H. Hubbell. 1973. The presence and possible significance of cross reactive antigens in TkUobTum-legurnl associations. Antonie van Leeuwenhoek J. Microbiol. Serol. 39:619-627. 63 -

PAGE 75

64 11 Clapp, C. E., and R. J. Davis. 1970. Properties of extracellular polysaccharides from Rfu.zobx.um. Soil Biol. Biochem. 2:109-117. 12. Dart, P. J., and F. V. Mercer. 1964. The legume rhizosphere. Arch. Mikrobiol. 47:344-378. 13 Dazzo, F. B., and D. H. Hubbell. 1974. A quantitative assay of insoluble polyvinylpyrrolidone. Plant Soil 40:435-440. 14 Dazzo, F. B., and D. H. Hubbell. 1975. Antigenic differences between infective and nomnfective strains of Rfuzobx-um tnJLbolll. Appl. Microbiol. 30:172-177. 15. Dazzo, F. B. , and D. H. Hubbell. 1975 Concanavalin A: lack of correlation between binding to Rfu.zobx.um and specificity in the Rfulzobxxim-legume symbiosis. Plant Soil 43: 717-722. 16 DeVay, J. E. 1975. Protein specificity in plant disease development: protein sharing between host and parasite. NATO Conference on Specificity in Plant Diseases, May 4-16, 1975, Sardinia Advanced Study Institute, Sardinia. 17 DeVay, J. E., W. C. Schnathorst, and M. S . _ Foda . 1966. Common antigens and host-parasite interactions. PJ1 J 328. In C Mirocha and I. Uritani (eds.), The dynamic role of molecular constituents in plant-parasitic interactions. Bruce Pub. Co., St. Paul, Minn. 18. Doubly, J. A., H. H. Flor, and C. O Clagett 1960. Relation of antigens of Mzlamptoiz ^"^^ lu^x.att a<aU&ilmum to resistance and susceptibility. Science 131:229. 19. Dudman, W. F. 1972. Detection of acidic P^sf ^arides in gels by DEAE-dextran. Anal. Biochem. 46-.668-6/J. 20. Fahraeus, G. 1957. The infection of clover root hairs by nodule bacteria studied by a simple glass slide technique. J. Gen. Microbiol. 16:374-381. 21. Goldman, M. 1968. Fluorescent antibody methods. Academic Press, N. Y. 303p. 22. Hamblin, J., and S. P. Kent. 1974 Possible role of phytohaemagglutinin in Vhataolu* vulgar L. Nature New Biology 245:28-30. 23. Hubbell, D. H. 1970. Studies on the root hair "curling factor" of Rhlzoblum. Bot. Gaz. 131.337-342. 24 Humphrey, B. A. 1959. Occurrence of 4-O-methyl glucuronic acid in Rfulzobxxxm gums. Nature (London) 134:1802.

PAGE 76

65 25. Hunter, W. 1973. Radioimmunoassay. In D. Weir (ed. ) , Handbook of experimental immunology, p. 17.1-17.33. Blackwell Scientific Pub. , Oxford, England. 26. Johnson, L. F., and E. A. Curl. 1972. Methods for research on the ecology of soil-borne plant pathogens, p. 92-96. Burgess Pub. Co., Minneapolis, Minn. 27. Jones, D. G. , and P. E. Russell. 1972. The application of immunofluorescence techniques to host plant/nodule bacteria selectivity experiments using T tuL^o LUim tizpznb . Soil Biol. Biochem. 4:277-282. 28. Rabat, E. A., and M. M. Mayer. 1961. Experimental immunochemistry, 2nd ed. Charles Thomas Pub. , Springfield, 111. 905p. 29. Li, D. , and D. H. Hubbell. 1969. Infection thread formation as a basis of nodulation specificity in Rh.izob.iu.mstrawberry clover associations. Can. J. Microbiol. 15:1133-1136. 30. Ljunggren, H. 1969. Mechanism and pattern of Rki.zobA.um invasion into leguminous root hairs. Physiol. Plant. Suppl. V. 82p. 31. Lov/ry, O. H., N. J. Rosebrough, A. L. Farr , and R. J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-269. 32. Lratky, A., P. Biely, and S. Bauer. 1975. Mechanism of 2-deoxy-D-glucose inhibition of cell wall polysaccharide and glycoprotein biosynthesis. Eur. J. Biochem. 54:459-467. 33. MacGregor, A. N. , and M. Alexander. 1972. Comparison of nodulating and non-nodulating strains of Rki.zob-lu.rn th.liolli. Plant Soil 36:129-139. 34. Markowitz, A. S. , S. H. Armstrong, and D. S. Kushner. 1960. Immunological relationships between the rat glomerulus and nephritogenic streptococci. Nature (London) 187:1095-1097. 35. Mowry, R. 1956. Alcian blue techniques for the histochemical study of acidic carbohydrates. J. Histochem. Cytochem. 4:470. 36. Munns, D. N. 1968. Nodulation of ttQ.di.cago i,at-iva in solution culture. I. Acid-sensitive steps. Plant Soil 28:129-146. 37. Napoli, C, F. Dazzo, and D. Hubbell. 1975. Ultrastructure of infection and "common antigen" relationships in Ae-i ckynomanz. 5th Australian Legume

PAGE 77

66 38. 39. 40. 41. Conference, March 18-21, 1975, Brisbane, Australia. Nanoli C A., F. B. Dazzo, and D. H. Hubbell. 1975. Production of cellulose microfibrils by RUzobAM*, Appl. Microbiol. 30:123-131. Naooli C N., and D. H. Hubbell. 1975. infrastructure of Kk'izoblum induced infection threads in clover root hairs Abstr. Ann. Meeting, Amer. Soc . Microbiol. April 27 May 2, 1975, New York. Nowotny, A. 1969. Basic exercises in immunochemistry . Springer-Verlag, N.Y. 197p. Pearlman, P., S. Hammarstrom, R. Lagercrantz, and B. E. GustSSn. 1965. Antigen from colon of germfree rats and antibodies in human ulcerative colitis. Annu. N. Y. Acad. Sci. 124:337-394. 42 Peters, R. J., and M. Alexander. 1966. Effect of legume exudates on the root nodule bacteria. Soil Science 102:380-387. 43 Pierce Chemical Company. 1972. Handbook of silylation. Pierce Chemical Company, Rockford, 111. 48 P . 44 Reporter, M. , D. Raveed , and G. Norris. 1975 Binding 44 ' IfTkTzo'bium' japon.cum to cultured soybean root cells: morphological evidence. Plant Science Letters 5.73 Schnathorst, W. C. 1973 Comparative antigenic structures of virulent and avxrulent isolates of Kantkomonat malv ac datum . p. 1330. Proc zna Congr. Plant Pathology, Amer. Phytopathol. Soc, Sept. 5-12, 1973, Minneapolis, Minn. Solheim, B., and J. Raa. 1973. ^araoterization of substances causing deformation o f root h^s °f &Uiolll. TulioUium .xzpnn& when inoculated with Kfc^oDx.um a. o j. Gen. Microbiol. 77:241-247. 47 Steel R. G. D. , and J. H. Torrie. I960 Principles * and procures if statistics. McGraw-Hill Book Co., Inc., New York. 4 81p. x iq£7 Thp ecoloav and etiology of human 48 disease! p. i06-i07 ChaflefThcas Pub. Co., Springfield, 111. 49. Vaugnt, R. , and A Bl eiweis -^1^4. Antigens of n Stfizvtococcu.6 mutant. II. maracceniaL , : resembling a glycerol teichoic acid in walls of strain BHT. Inf. Immun. 9:60-67. 45. 46,

PAGE 78

67 50. Vincent, J. M. 1970. A manual for the practical study of root nodule bacteria. IBP Handbook No. 15. Blackwell Scientific Pub., Oxford, England. 164p. 51. Weaver, R. , and L. Frederick. 1972. A new technique for most-probable-number counts of RlUzobZam. Plant Soil 36:219-222. 52. Wimalajeewa, D. L. S. , and J. E. DeVay. 1971. The occurrence and characterization of a common antigen relationship betwen Ui,tilciQO maydii and Zea maij-6. Physiol. PI. Path. 1:523-535. 53. Worthington Biochemical Corporation. 1972. Worthington enzyme manual. Worthington Biochemical Corporation, Freehold, N. J. 216p. 54. Yao, P. Y., and J. M. Vincent. 1969. Host specificity in the root hair "curling factor" of Rhizobiixm spp. Aust. J. Biol. Sci. 22:413-423. 55. Zabriskie, J. B. 1967. Mimetic relationships between group A streptococci and mammalian tissue. Adv. Immunol. 7:147-188.

PAGE 79

BIOGRAPHICAL SKETCH Frank Bryan Dazzo was born April 8, 1948, in Miami, Florida. He was graduated from Coral Gables High School in Coral Gables, Florida, in June, 1966. In June, 1968, he received an Associate of Science degree from Young Harris College, Young Harris, Georgia, where he was elected to Who's Who in American Junior Colleges. In August, 197 0, he received his Bachelor of Science degree with a major in bacteriology from The Florida State University, Tallahassee, Florida. He began his graduate studies at the University of Florida in September, 197 0, and received his Master of Science degree in microbiology there in June, 197 2. In November, 1974, he was the recipient of the President's Award from the Southeastern Branch of the American Society for Microbiology for outstanding research presented at that meeting In December, 1974, he was awarded a Grant-in-Aid for Research by Sigma Xi, the Scientific Research Society of North America. m is a member of the American Society for Microbiology, the Southeastern Branch of the American Society for Microbiology, the Florida Soil and Crop Science Society, the Society of Sigma Xi, and the National Honorary Society of Phi Kappa Phi. He is currently a candidate for the Ph. D. degree in the Department of Microbiology, University of Florida. He is married to Olgalina G. Dazzo. 68 -

PAGE 80

I certify that "I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. David H. HufcSbell/ Chai'rman Associate Professor of Soil Microbiology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Paul H. Smith Professor of Microbiology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Arnold S. Bleiweis Associate Professor of Microbiology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. I* Edward M". Hoffm Associate Profe wflUV^ Microbiology

PAGE 81

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. James" F 1 ". Preston, III Associate Professor of Microbiology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Raghavan Charudattan Assistant Professor of Plant Pathology This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December, 197 5 ck /