Modulation of Salmonella infection by the lectins of Canavalia ensiformis (Con A) and Galanthus nivalis (GNA) in a rat model in vivo

Modulation of Salmonella infection by the lectins of Canavaliaensiformis (Con A) and Galanthus nivalis (GNA) in a rat modelin vivo

P.J. Naughton*, G. Grant, S. Bardocz and A. PusztaiRowett Research Institute, Aberdeen, UK

46/11/99: received 5 November 1999, revised 31 January 2000 and accepted 2 February 2000

P .J . N A U G H T O N , G . GR A N T , S . B A R D O C Z A N D A . P U S Z T A I . 2000. The plant lectins,

Concanavalin A (Con A) and Galanthus nivalis agglutinin (GNA) have been prefed to rats for

3 d pre- and 6 d postinfection with Salmonella typhimurium S986 or Salm. enteritidis 857. Con

A signi®cantly increased numbers of Salm. typhimurium S986 in the large intestine and in

faeces, and severely impaired growth of the rats, more severely than is the case of infection

with Salmonella typhimurium alone. Con A had much less effect on rats infected with Salm.

enteritidis 857 only showing a signi®cant increase in numbers in the colon, accompanied by

intermittent increases of Salmonella in the faeces during the study. GNA signi®cantly

reduced pathogen numbers in the lower part of the small bowel and the large intestine of rats

infected with Salm. typhimurium S986 and signi®cantly improved rat growth. GNA had little

effect on infection by Salm. enteritidis 857 with slight decreases in Salmonella numbers in the

small intestine and large intestine and transient increases in the faeces.

INTRODUCTION

Adherence is probably an important factor in bacterial colo-

nization (Melhem and LeVerde 1984), and numerous sur-

face carbohydrates of various forms on animal cells provide

a plethora of attachment points for bacterial lectins

(Karlsson et al. 1992; Sharon 1993). Cell surface-bound car-

bohydrate-binding proteins (animal lectins) may also act as

attachment points for micro-organisms carrying the appro-

priate sugars on their surface (Sheriff et al. 1994). This

mechanism is used by macrophages for nonopsonic phago-

cytosis for clearance of micro-organisms in tissues (Ofek

et al. 1992). Cell surface lectins (Sharon and Lis 1989) are

also involved in cell-to-cell attachment and/or signalling.

Thus, lectins could be involved in the binding of bacteria to

target cells by a variety of mechanisms. Numerous compo-

nents have been proposed for inhibiting or promoting bac-

terial binding in the gastrointestinal (GI) tract. Various

sugars (Allen et al. 1997), bacterial antigens (Thorns 1995),

probiotics (MacFarlane and Cummings 1999) and synbiotics

(Collins and Gibson 1999) have been investigated.

Plant lectins can bind to the GI tract in vivo and induce

changes in the gut structure, glycosylation and function.

Therefore, they have the potential to in¯uence adherence

and gut colonization by pathogens. Beuth et al. (1995) high-

lighted the importance of lectins in the prevention of bacter-

ial infections. Work by Giannasca et al. (1996) suggests that

one receptor exploited by Salmonella on Caco-2 cells is the

Gal-b(1±3) GalNAc structure which is also recognized by

peanut agglutinin (PNA). Consequently, PNA has been

shown to inhibit Salmonella binding to Caco-2 cells. More

recently Poschet and Fairclough (1999) have shown that a

number of lectins can prevent the invasion of Caco-2 cells

by Salmonella in vitro. Lectins may also interact directly

with bacteria and can be been used for typing bacterial

strains (Kellens et al. 1995) and for the targeted separation

of Enterobacteriaceae in vitro (Porter et al. 1998).

The lectin GNA from the snowdrop (Galanthus nivalis)

bulb is speci®c for terminal a-1-3-linked mannose. Orally

administered GNA inhibited kidney bean lectin-induced

Escherichia coli growth in the rat small intestine in vivo

(Pusztai et al. 1990; Pusztai et al. 1993). GNA has, how-

ever, little or no effect on intestinal metabolism itself

(Pusztai et al. 1995) and may therefore have uses in pre-

venting or interfering with colonization by pathogenic

micro-organisms.

Seeds of Jack bean (Canavalia ensiformis) contain

Concanavalin A (Con A), a mannose/glucose binding lec-

Correspondence to: P.J. Naughton, Danish Institute of Agricultural Sciences,

Research Centre Foulum, PO Box 50, DK-8830, Tjele, Denmark (e-mail:

[email protected]).

*Present address: Department of Animal Nutrition and Physiology, Danish

Institute of Agricultural Sciences, Research Centre Foulum, PO Box 50, DK-

8830, Tjele, Denmark.

Journal of Applied Microbiology 2000, 88, 720ÿ727

= 2000 The Society for Applied Microbiology

tin. Con A can induce or inhibit cell-mediated lympholysis

in murine lymphocytes in vitro (Sitkovsky et al. 1982) and

can activate macrophages in vitro (Maldonado et al. 1994).

Con A facilitates the clearance of Candida albicans by

macrophages from tissues (Felipe et al. 1995). Summer and

Howell (1936) showed that Con A agglutinated bacteria

from the genera Mycobacterium and Actinomyces, a prop-

erty of the lectin which has been repeatedly con®rmed

(Pistole 1981). Con A also binds to viruses such as

Newcastle disease V4 strain (Rehmani and Spradbrow

1995) and has been shown to inhibit attachment of Salm.pullorum to chicken gut (Zhou et al. 1995) and to block

attachment of Giardia lamblia trophozoites to Caco-2 cells

(Katelaris et al. 1995). However, Con A has been found to

promote adherence of Salmonella to isolated viable intest-

inal cells and to promote adherence to intestinal loops

(Abud et al. 1989). The lectin might therefore promote

bacterial adherence to the small intestine. The effects of

dietary GNA and Con A on Salmonella infection have now

been studied in vivo using a rat model.

MATERIALS AND METHODS

Bacterial strains

Salm. enteritidis 857 phage type 4 (Van Asten et al. 1995)

and Salm. typhimurium S986 phage type 49 (Naughton et al.1996) were used in this study. These wild type strains were

grown under static conditions at 37 �C for 48 h in nutrient

broth (CM1, Unipath, Basingstoke, UK) (for studies invivo).

Lectins

Concanavalin A (speci®city D-mannnose/glucose) was

obtained from Sigma (St Louis, MO, USA). A strictly

mannose-speci®c lectin, GNA, was isolated from snowdrop

(Galanthus nivalis) bulbs by af®nity chromatography on

mannose-agarose (Van Damme et al. 1987).

Lectin aggregation of Salmonella in vitro

This method was based on the procedure of Abud et al.(1989). Bacterial cells were grown without shaking for 48 h

in nutrient broth (CM1; Unipath) at 37 �C. Bacterial cells

were collected by centrifugation (1300 g, 10 min), washed

twice and resuspended in PBS (pH 7�4). The absorbance of

the suspensions as measured at 620 nm in a spectrophot-

ometer (Varian, Mountain View, CA, USA; DMS 200 u.v.visible spectrophotometer), and the suspensions were

adjusted to an A620 equivalent of 5�0. Lectin solutions

were prepared at a concentration of 1 mg mlÿ1 (w/v) in

PBS (pH 7�4). Agglutination assays were carried out in

glass tubes (12 by 75 mm). Assays consisted of 10 ml lectin

and 90 ml cell suspension. A negative control consisted of

10 ml buffer instead of lectin solution. Sugar inhibition tests

were carried out with 2 mol lÿ1D-glucose and D-mannose.

Tubes were left at room temperature (20 �C) for 1 h and

agglutination was viewed under the microscope.

The degree of aggregation was estimated visually as fol-

lows:

3�: 80±100% of cells clumped (under a microscope)



tr:< 10% of cells clumped (under a microscope)

Animal handling

Newly weaned 19 d-old male Hooded Lister-speci®c patho-

gen-free rats of the Rowett colony were housed singly in

metabolism cages and fed a good quality semisynthetic diet

based on lactalbumin (100 g kgÿ1 diet) as the sole protein

source (Grant et al. 1995). Food was freely available for 12

d. Once the rats reached 80 � 1 g, they were transferred to

a ¯exi-®lm isolator (Moredun Animal Health, Edinburgh,

UK) were kept singly in metabolism cages and fed the

above diet at 6 g/rat dÿ1 (given as two feeds over the day)

for 4 days. Water was available ad libitum.

Experimental procedures in vivo

The small animal unit of the Rowett Research Institute is

licensed under the UK Animals (Scienti®c Procedures) Act

1986, and its operation is regulated by both the animal wel-

fare unit of the institute and the appropriate government

inspectorate. All management and experimental procedures

conducted during this study were done in strict accordance

with the requirements of the UK Animals (Scienti®c

Procedures) Act 1986 by staff personally licensed under

this act to perform such procedures.

Prior to infection with either Salm. typhimurium or Salm.enteritidis, the rats were prefed for 3 d either Con A or

GNA incorporated into the control lactalbumin diet or

control diet without lectin. The diet was given as two feeds

of 3�3 g diet over each day. On day one the rats received

10 mg lectins and on days two and three, the rats received

20 mg lectin dÿ1 in their diet. Overnight the lectin treated

groups received lectin (2 mg mlÿ1) in their drinking water.

Control rats were given water without lectin. Control

groups treated with Con A alone and GNA alone were

included.

On the eve of the fourth day rats were fasted to reduce

the effects of stomach pH and then given either 1ml saline

(controls) or 1 ml saline containing 10 mg lectin by intra-

gastric intubation. Two hours later they were given 1 ml

broth of either Salm. enteritidis 857 or Salm. typhimurium

721P LA N T L E C T I N S A L T E R S A L M O N E L L A I N F E C T I O N

= 2000 The Society for Applied Microbiology, Journal of Applied Microbiology, 88, 720ÿ727

S986 (1� 108 CFU/mlÿ1) by gavage. A further 1�5 h later

the test group was fed 3�3 g control diet containing 20 mg

lectin while the control group received 3�3 g control diet

alone. This feeding protocol was repeated in the evening

and for the remaining 5 d of the experiment. Lectin

(2 mg mlÿ1) were also available in the water throughout the

experiment for the lectin-treated groups. On day 6, the rats

were fed 1�5 g of diet with or without 10 mg of lectin. Postmortems were performed exactly 2 h later and the animals

dissected. Tissues and organs were removed aseptically for

the enumeration of bacteria.

Bacteriological examination

The tissues were weighed and homogenized in maximum

recovery diluent (Unipath; CM733). A 10-fold dilution ser-

ies was prepared for the homogenates to a ®nal dilution of

1� 10ÿ7. Salmonella were enumerated on XLD agar

(Unipath; CM469) after incubation at 37 �C for 16 h.

Rogosa agar (Unipath; CM627) was used for the enumera-

tion of Lactobacilli, and a selective medium consisting of

Columbia agar (Unipath; CM331) and 0�5% (v/w) propio-

nic acid (Fisons, Loughborough, UK) was used for the iso-

lation of Bi®dobacteria. These cultures were incubated for

48 h at 37 �C in an anaerobic cabinet (Don Whitley

Scienti®c, Shipley, UK) and colonies were counted (Miles

and Misra 1938). The presence of Salmonella was con-

®rmed by speci®c antisera (Murex, Berkshire, UK) and

API 20E (BioMerieux, Marcy l'Etoile, France). Lactobacilli

and Bi®dobacteria were presumptively identi®ed by Gram-

stain, their ability to grow on selective media, colonial mor-

phology and catalase reaction. The detection limit was 500

cfu gÿ1 wet weight or per ml. Spleen and liver samples

were plated out according to the pour-plate method

(Collins and Lyne 1995); the detection limit was 10 cfu gÿ1

wet weight.

Statistics

Data were analysed using one-way analysis of variance by

the Minitab computer program (Minitab, State College,

PA, USA), and multiple comparisons were done by the

Tukey test using Instat statistical package (Graphpad

Software, San Diego, CA, USA).

RESULTS

Lectin aggregation of Salmonella in vitro

Two lectins, Con A and GNA were tested for their ability

to aggregate Salmonella strains (Table 1). Most Salm. typhi-murium strains we have tested were aggregated by Con A.

In contrast, there was virtually no aggregation of these

Salm. typhimurium strains by GNA. The Con A-mediated

aggregation was due to the speci®c carbohydrate binding

properties as binding of lectin to the bacteria was prevented

by sugars. Aggregation of Salm. enteritidis occurred in a

majority of strains examined in the presence of GNA (5/6)

and Con A (4/6).

Effects of Con A in vivo

Apart from the colon there were no signi®cant increases in

the Salm. enteritidis 857 population in the GI tract of

infected rats as a result of the inclusion of Con A in the

diet (Table 2). Con A did not alter the Lactobacilli or

Bi®dobacteria populations (Lactobacilli (log10counts per

gram: jejunum 3�1 � 0�5, ileum 4�8 � 1�2, caecum 6�3 � 0�4,colon 6�0 � 0�8. Bi®dobacteria (log10 counts gÿ1): jejunum

2�8 � 0�2, ileum 3�0 � 0�7, caecum 4�3� 1�3, colon 4�0 � 2�4)in noninfected control rats. Dietary Con A had a marked

effect on Salm. typhimurium S986 infection (Table 2). Most

tissues had increased numbers of Salmonella with differ-

ences in the numbers present in the caecum, colon and

liver reaching signi®cance (P< 0�05).

Effects of GNA in vivo

Inclusion of GNA in diets of infected rats did not signi®-

cantly alter the numbers of Salm. enteritidis 857 in the gut

or internal organs (Table 2). In contrast, GNA-treatment

reduced numbers of Salm. typhimurium S986 in the ileum

with levels reaching signi®cance (P< 0�05) in the caecum

and colon of infected rats (Table 2).

Effects of lectins on Salmonella numbers in the

faeces

Control rats infected with Salm. typhimurium or Salm.enteritidis excreted high amounts of the pathogen in their

faeces through the 6-d trial (Table 3). In both cases, faecal

excretion of the bacteria tended to be further increased

when Con A was included in the diet. Dietary GNA also

tended to increase faecal excretion of Salm. enteritidis but

appeared to have only transient effects on faecal outputs of

Salm. typhimurium.

Weight gain

Rats infected with Salm. typhimurium S986 grew more

slowly than noninfected controls (Fig. 1). This growth

retardation was exacerbated when Con A was included in

the diet of the infected rats. Indeed these rats grew little

over the 6 d. In contrast, the growth rates of rats infected

with Salm. typhimurium S986 were similar to those of non-

infected controls when GNA was added to the diet. Salm.

722 P . J . N A UG H T O N E T A L .


Table 2 Viable Salmonella counts detected in tissues after 6 d

Jejunum Ileum Caecum Colon Spleen Liver MLN*

Salm. enteritidis 857 (SE{)SE�LA 3�1 � 0�5 4�9 � 1�3 5�3 � 0�5 3�9 � 1�7a 3�0 � 0�4 2�9 � 0�4 4�8 � 0�3SE�GNA 3�5 � 0�7 5�6 � 0�3 4�6 � 1�8 5�1 � 1�4ab 3�4 � 0�7 2�9 � 0�5 5�2 � 0�2SE�Con A 3�9 � 0�8 5�3 � 0�2 5�7 � 0�6 5�1 � 0�7b 3�2 � 0�2 2�7 � 0�6 5�0 � 0�2

Salm. typhimurium S986 (STM{)STM�LA 5�8 � 4�7 6�8 � 0�3a 8�8 � 0�9a 8�4 � 0�8a 4�1 � 0�6 4�1 � 0�6 5�9 � 0�4STM�GNA 4�7 � 0�3 6�4 � 0�1b 7�1 � 0�8b 6�5 � 1�6b 3�7 � 1�9 3�6 � 1�9 5�8 � 0�1STM�LA 4�5 � 1�1 5�9 � 1�0 7�1 � 1�6a 6�3 � 1�7a 3�6 � 0�7 3�3 � 0�6a 5�8 � 0�2STM�Con A 5�4 � 0�3 6�8 � 0�4 8�8 � 0�3b 8�9 � 0�4b 4�3 � 0�4 4�3 � 0�6b 6�3 � 0�6

Values are means, log cfu gÿ1 faeces (with S.D.). Groups (n� 5) of rats were gavaged with either Salm. enteritidis 857 or Salm. typhimurium

S986 and fed on different diets: LA ± control lactalbumin for 6 d; Con A or GNA were included in the diet for 3 d pre- and 6 d

postinfection. For Salm. typhimurium each lectin treatment has a control. For each experimental group mean values with different

superscripts in vertical columns are signi®cantly different (P< 0�05).*Mesenteric lymph nodes.

{SE±Salm. enteritidis.

{STM±Salm. typhimurium.a,b represent signi®cance at P< 0�05.

Table 1 Aggregation of Salmonella by lectins

Strain Serovar Source GNA Con A

Wharton pt104* Salm. typhimurium Chicken tr 3�2 Acres pt20* Salm. typhimurium Bovine tr tr

Gloucester pt1* Salm. typhimurium Human tr tr

Swindon pt32* Salm. typhimurium Human tr 3�S986 pt49* Salm. typhimurium Chicken tr 3�F98* Salm. typhimurium Chicken tr 3�195* Salm. cholerasuis Procine tr tr

76 Sa 88{ Salm. enteritidis Chicken egg 2� 2�The Neale* Salm. typhimurium Chicken tr 2�A50* Salm. cholerasuis Procine tr 3�181 Sa 91{ Salm. typhimurium Bovine tr 2�Bangor 0T44* Salm. typhimurium Bovine tr 1�p132344* Salm. enteritidis Chicken 2� 2�p125589* Salm. enteritidis Chicken 2� 2�Beau ville* Salm. typhimurium Chicken tr 2�S1339* Salm. dublin Bovine tr tr

1116* Salm. typhimurium Chicken tr 2�LA5{ Salm. enteritidis Chicken 3� tr

Cambridge* Salm. typhimurium Turkey tr 3�857{ Salm. enteritidis Chicken 3� tr

W118x Salm. typhimurium Human tr 3�SAL33�{ Salm. enteritidis Human tr 3�TMLx Salm. typhimurium Human tr 3�

*P. Barrow, Institute of Animal Health, Compton, UK.

{P. Lintermans, National Institute of Veterinary Research, Brussels, Belgium.

{C.J. Thorns, Central Veterinary Laboratory, UK

xJ. Stephen, University of Birmingham, UK.

{M. Loos, Institute of Medical Microbiology, Johannes Gutenberg-University, Mainz, Germany.



enteritidis 857 impaired the growth of rats (Fig. 1). This

was not affected by addition of Con A to the diet.

However, GNA did cause a slight improvement in weight

gain of rats infected with Salm. enteritidis 857.

Con A caused a slight (approximately 11%) but not sta-

tistically signi®cant reduction in weight gains when fed to

noninfected control rats (Fig. 1). GNA did not alter rat

growth.

DISCUSSION

Adherence of bacteria to epithelial cells appears to be a key

initial step in infection of the gut by a pathogen (Ofek and

Sharon 1988). This attachment of the bacteria to entero-

cytes is primarily mediated by the ®mbrial adhesins/lectins

which recognize and bind to speci®c carbohydrate struc-

tures on the epithelial surface. Plant lectins, with the same

carbohydrate speci®city as bacterial lectins, can also bind to

these carbohydrate structures on the gut epithelium. Thus,

if they are present in the gut lumen in excess, they have

the potential to modulate colonization by micro-organisms.

Thus, we have shown that GNA reduces Salm. typhimur-ium numbers in vivo, a property previously shown for

GNA but in a Chlamydia trachomatis infection in vitro(Amin et al. 1995). Con A has been shown to increase

Salmonella numbers in the rat in situ (Abud et al. 1989).

We have now shown that Con A increases Salmonella num-

bers in the rat in vivo.

Aggregation of Salmonella by lectins in vitro

Lectin±bacterial interaction studies have concentrated on

the use of lectins as a means of characterizing bacterial

populations or bacterial structures. Plant lectins are used

routinely for mapping glycosylation patterns of mammalian

tissues (Pusztai et al. 1995). Basu et al. (1994) used the sia-

lic acid-binding lectin from Cepaea hortensis as a means of

characterization of Salm. djakarta and Salm. isaszeg. Thus,

it was not unexpected that we found certain lectins were

capable of causing aggregation of Salmonella. The extent of

Table 3 Viable Salmonella counts obtained in faeces samples taken on different days

Day

1 2 3 4 5 6

Salm. enteritidis (SE*)

SE�LA 7�5 � 0�6 5�8 � 0�8 5�2 � 0�6a 5�7 � 0�9 5�1 � 0�8a 4�6 � 1�4a

SE�GNA 5�7 � 1�8 6�8 � 1�0 7�6 � 0�5b 6�6 � 1�3 6�7 � 0�5b 6�9 � 1�3b

SE�Con A 7�4 � 0�5 6�8 � 1�1 6�4 � 0�5b 7�0 � 1�0 6�5 � 0�8b 6�7 � 0�8b

Salm. typhimurium (STM{)STM�LA 7�7 � 1�8 6�1 � 0�7a 7�9 � 1�1 7�1 � 1�2 8�1 � 0�9 7�9 � 0�9STM�GNA 7�7 � 1�7 7�5 � 0�8b 7�6 � 1�1 6�6 � 2�5 7�7 � 1�6 7�5 � 1�5STM�LA 5�4 � 0�6a 5�9 � 0�8a 5�9 � 0�6a 7�1 � 1�2 6�8 � 1�4 7�7 � 1�8a

STM� ConA 8�3 � 1�4b 7�8 � 0�8b 8�6 � 1�0b 7�9 � 1�5 7�8 � 1�3 9�8 � 0�6b

Values are means, log cfu gÿ1 faeces (with S.D.). Groups (n� 5) of rats were gavaged with either Salm. enteritidis 857 or Salm. typhimurium

S986 and fed on different diets: LA ± control lactalbumin for 6 d; Con A or GNA were included in the diet for 3 d pre- and 6 d

postinfection. For Salm. typhimurium each lectin treatment has a control. For each experimental group mean values with different

superscripts in vertical columns are signi®cantly different (P< 0�05).*SE±Salm. enteritidis.

{STM±Salm. typhimurium.a,b represent signi®cance at P< 0�05.

Fig. 1 Effects of lectins alone in the diet and in the presence of

Salmonella on the weight gain of rats 6 d postinfection. Data are

expressed as mean (n� 5) with S.D. of mean

724 P . J . N A UG H T O N E T A L .


the aggregation was highly dependent on the speci®city of

the lectin used. We found that GNA did not aggregate the

Salm. typhimurium strains tested in this study but did bind

to Salm. enteritidis strains (Table 1). In contrast, Con A

caused the aggregation of most, but not all strains of Salm.typhimurium and Salm. enteritidis tested in this study.

Con A, which is mannose/glucose speci®c, aggregates a

variety of Gram-negative bacteria especially Salmonellaspp. (Le Minor et al. 1973). The sites of Con A-binding

are thought to be the exposed sugars of the bacterial lipo-

polysaccharide and the O-antigen factor 1 (Liener 1976).

Early studies on Con A reactivity showed that the lectin

bound to lipopolysaccharide from some Enterobacteriaceae

but not that from others (Doyle et al. 1968). Since Con A

aggregated Salmonella strains and was also found to

increase the attachment of Salm. typhimurium to gut epithe-

lium in situ (Abud et al. 1989), the effects of Con A on gas-

trointestinal infection merited study.

Effect of lectins on rat growth

Con A (20 mg ratÿ1 dÿ1) caused a slight reduction in weight

gain when it was fed to noninfected rats whereas GNA

(20 mg ratÿ1 dÿ1) had little effect on growth. This was con-

sistent with earlier ®ndings with higher dietary intakes of

these lectins (Oliveira et al. 1994; Pusztai et al. 1996).

Modulation of Salmonella infection in vivo

Con A appeared to increase the severity of infection in rats

dosed with strain Salm. typhimurium S986 but did not alter

the numbers of Salm. enteritidis 857. Con A is highly resis-

tant to proteolysis in vivo and is stable during its passage

through the gastrointestinal tract (Nakata and Kimura

1985). At high levels in the diet it disturbs brush border

membrane function and interferes with reformation of

brush border membranes (Nakata and Kimura 1986).

There may also be some cytotoxic effects (Lorenz-Meyer

et al. 1985), particularly in young animals (Weaver and

Bailey 1987). The lectin binds to brush borders in vitro(Etzler and Branstrator 1974) and in vivo and thus leads to

increased shedding of brush border membranes, accelerated

cell loss and slight villus atrophy (Lorenzsonn and Olsen

1982). These changes require the presence of bacteria in

the gut since Con A has been shown to have no effect on

the internal structure and function of germ-free animals

(Pusztai et al. 1995).

Some polyvalent lectins can recognize receptors on bac-

teria as well as on the enterocyte, such as mannose residues

located on the O-antigen of Salm. typhimurium and glyco-

proteins on the enterocyte (Alexander 1981). For example,

a®mbriate bacteria were shown to bind to Con A-agarose

beads in vitro (Abud et al. 1989). Bar-Shavit and Goldman

(1976) have shown Con A-mediated attachment and inges-

tion of yeasts cells by macrophages. Gallily et al. (1984),

studied the wheat germ agglutinin-mediated uptake of bac-

teria by murine peritoneal macrophages. Con A may there-

fore form a bridge between the brush border and the

bacterium and thus promote bacterial adherence. The

observation that the presence of Con A in the diet did not

alter numbers of lactobacilli or bi®dobacteria populations

suggests that the increase in Salm. typhimurium S986 num-

bers may be speci®c, at least to this serotype.

Con A can stimulate histamine release and activation of

epidermal growth factor (EGF). Con A binds to the EGF

receptor on mammalian cells (Carpenter and Cohen 1977)

and inhibits the dimerization and tyrosine phosphorylation

of the receptor leading to inhibition of mitogenicity of

EGF (Hazan et al. 1995). Salm. typhimurium has similar

effects on many of these pathways (GalaÂn et al. 1992). Con

A and Salm. typhimurium S986 by acting together may

have a synergistic effect on the mechanism by which colo-

nization and invasion take place and thereby increase infec-

tivity. This is borne out by a further signi®cant reduction

in the weight gain of rats infected with Salm. typhimuriumS986 by giving Con A in the diet (Fig. 1).

Previous studies in this laboratory have shown that the

incorporation of GNA alone in the diet of rats at

42 mg ratÿ1 dÿ1 for 6 d did not alter the numbers of lactose

fermenters, non-lactose fermenters or lactobacilli from LA

control levels (Pusztai et al. 1993). Far from having a dele-

terious effect, GNA reduced Salm. typhimurium S986 num-

bers in the ileum, caecum and colon (Table 2). GNA has a

selective speci®city for terminal mannose structures. High-

mannose glycans are ubiquitous components of bacterial

pathogens, parasites, yeasts and some viruses (Ezekowitz

and Stahl 1988), and are also present in many other cells

particularly immature or damaged cells. It is therefore pos-

sible that GNA can agglutinate Salmonella in vivo and in so

doing, prevent attachment to the substratum. However, we

have shown in the case of Salm. typhimurium S986 that

GNA does not aggregate the bacterium in vitro (Table 1)

whereas it will aggregate Salm. enteritidis 857. Therefore

the ability of GNA to decrease the numbers of Salm. typhi-murium S986 is unlikely to be due to the aggregation of

bacteria. It is possible that GNA may alter cellular metabo-

lism and reduce the level of expression of enterocyte mem-

brane glyconjugates essential for attachment of Salm.typhimurium S986.

Salm. typhimurium S986 causes major changes to gut

epithelium. In particular it triggers high rates of epithelial

cell division and turnover which can then facilitate infec-

tion by the bacteria (Naughton et al. 1995). GNA may

bind to these mannose-rich sites on enterocytes and thus

prevent association of the bacterium, alternatively GNA

may slow down the rate of cell turnover and limit the



damaging effects of Salm. typhimurium S986 by reducing

the numbers of immature cells reaching the villus surface.

CONCLUSIONS

The lectins studied had different effects on infection by

Salmonella. Con A appeared to promote adherence by

Salm. typhimurium S986 whereas GNA reduced adherence

of the pathogen. This suggests that lectins are likely to be

selective in their effects on the gut ¯ora. Speci®c lectins

might be used to promote or suppress particular bacterial

types within the gut ¯ora and thereby improve the health

of the host.

ACKNOWLEDGEMENTS

This work was supported in part by the Scottish Of®ce

Agriculture Environment and Fisheries Department and by

a doctoral bursary to PN from the European Commission

(Cont. No. AGRF-CT93±5408) and forms part of COST

98.

REFERENCES

Abud, R.L., Lindquist, B.L., Ernst, R.K., Merrick, J.M.,

Lebenthal, E. and Lee, P.C. (1989) Concanavalin A promotes

adherence of Salmonella typhimurium to small intestinal mucosa

of rats. Proceedings the Society for Experimental Biology and

Medicine 192, 81±86.

Alexander, M. (1981) Why microbial predators and parasites do

not eliminate their prey and hosts. Annual Review of

Microbiology 35, 113±133.

Allen, V.M., Fernandez, F. and Hinton, M.H. (1997) Evaluation

of the in¯uence of supplementing the diet with mannose or

palm kernel meal on Salmonella colonisation in poultry. British

Poultry Science 38, 485±488.

Amin, K., Beillevarie, D., Mahmoud, E., Hammar, L., Mardh,

P.-A. and Froman, G. (1995) Binding of Galanthus nivalis lectin

to Chlamydia trachomatis and inhibition of vitro infection. Acta

Pathologica, Microbiologica et Immunologica Scandinavica 103,

714±720.

Bar-Shavit, Z. and Goldman, R. (1976) Concanavalin A-mediated

attachment and ingestion of yeast cells by macrophages.

Experimental Cell Research 99, 221±236.

Basu, S., Schlecht, S., Wagner, M. and Mayer, H.L. (1994) The

sialic acid-containing lipopolysaccharides of Salmonella djakarta

and Salmonella isaszeg (serogroup O:48): Chemical characteriza-

tion and reactivity with the sialic acid-binding lectin from

Cepaea hortensis. FEMS Immunolology and Medical Microbiology

9, 189±198.

Beuth, J., Ko, H.L., Pulverer, G., Uhlenbruck, G. and

Pichlmaier, H. (1995) Importance of lectins for the prevention

of bacterial infections and cancer metastases. Glycoconjugate

Journal 12, 1±6.

Carpenter, G. and Cohen, S. (1977) In¯uence of lectins on the

binding of 125I-labelled EGF to human ®broblasts.

Biochemistry Biophysics Research Communications 79, 545±552.

Collins, M.D. and Gibson, G.R. (1999) Probiotics, prebiotics, and

synbiotics: approaches for modulating the microbial ecology of

the gut. American Journal of Clinical Nutrition 69, 1052S±

1057S.

Collins, C.H. and Lyne, P.M. (1995) Microbiological Methods. p.

834. Oxford: Butterworth-Heinemann Ltd.

Doyle, R.J., Woodside, E.E. and Fishel, C.W. (1968) Proton±

polyelectrolyte interactions. The concanavalin A precipitin reac-

tion with polyelectrolytes and polysaccaride derivatives.

Biochemistry Journal 106, 35±40.

Etzler, M.E. and Branstrator, M.L. (1974) Differential localiza-

tion of cell surface and secretory components in rat intestinal

epithelium by use of lectins. Journal of Cell Biology 62, 329±

343.

Ezekowitz, R.A.B. and Stahl, P.D. (1988) The structure and func-

tion of vertebrate mannose lectin-like proteins. Journal of Cell

Science 9, 121±133.

Felipe, I., Bim, S. and Somensi, C.C. (1995) Increased clearance

of Candida albicans from the peritoneal-cavity of mice pre-

treated with Concanavalin A or jacalin. Brazilian Journal of

Medical Biology Research 28, 477±483.

GalaÂn, J.E., Pace, J. and Hayman, M.J. (1992) Involvement of the

epidermal growth factor receptor in the invasion of cultured

mammalian cells by Salmonella typhimurium. Nature 357, 588±

589.

Gallily, R., Vary, B., Stain, I. and Sharon, N. (1984) Wheat germ

agglutinin potentiates uptake of bacteria by murine peritoneal

macrophages. Immunology 52, 679±686.

Giannasca, K.T., Giannasca, P.J. and Neutra, M. (1996)

Adherence of Salmonella typhimurium to Caco-2 Cells:

Identi®cation of a glycoconjugate receptor. Infection and

Immunity 64, 135±145.

Grant, G., Doward, P.M., Buchan, W.C., Armour, J.C. and

Pusztai, A. (1995) Consumption of diets containing raw soya-

beans (Phaseolus vulgaris) cowpeas (Vigna unguiculata) or lupin

seeds (Lupininus angustifolius) by rats for up to 700 days: effects

upon body composition and organ weights. British Journal of

Nutrition 73, 17±29.

Hazan, R., Krushel, L. and Crossing, K.L. (1995) EGF receptor-

mediated signals are differentially modulated by Concanavalin

A. Journal of Cellular Physiology 162, 74±85.

Karlsson, K.A., AÊ ngstroÈm, J., BergstroÈm, J. and Lanne, B. (1992)

Microbial interaction with cell surface. Acta Pathologica,

Microbiologica et Immunologica Scandinavica 100, 71±83.

Katelaris, P.H., Naeem, A. and Farthing, M.J.G. (1995)

Attachment of Giardia lamblia trophozoites to a cultured

human intestinal-cell line. Gut 37, 512±518.

Kellens, J.T., Jacobs, J.A., Peumans, W.J. and Stobberingh, E.E.

(1995) Agglutination of Streptococcus milleri by lectins. Journal

of Medical Microbiology 41, 14±19.

Le Minor, L., Tournier, P. and Chalon, A.M. (1973) The aggluti-

nation of Concanavalin A of certain Gram negative bacilli; a

study into the relationship with the somatic antigens of

Salmonella serotypes. Annals in Microbiology 124A, 467±476.

726 P . J . N A UG H T O N E T A L .


Liener, I.E. (1976) Phytohemagglutinins. Annual Review of Plant

Physiology 27, 291±319.

Lorenz-Meyer, H., Roth, P., Elsasser, P. and Hahn, U. (1985)

Cytotoxicity of lectins on rat intestinal mucosa is enhanced by

neuraminidase. European Journal of Clinical Investigation 15,

227±234.

Lorenzsonn, V. and Olsen, W.A. (1982) In vivo responses of rat

intestinal epithelium to intraluminal dietary lectins.

Gastroenterology 82, 838±848.

Maldonado, G., Porras, F., Fernandez, L., Vazquez, L. and

Zenteno, E. (1994) Effect of lectins on mouse peritoneal macro-

phage phagocytic-activity. Immunology Investigation 23, 429±

436.

MacFarlane, G.T. and Cummings, J.H. (1999) Probiotics and

prebiotics: can regulating the activities of intestinal bacteria

bene®t health? British Medical Journal 318, 999±1003.

Melhem, R.F. and LeVerde, P.T. (1984) Mechanism of interac-

tion of Salmonella and Schistoma species. Infection and Immunity

44, 274±281.

Miles, A.A. and Misra, S.S. (1938) The estimation of the bacterial

power of blood. Journal of Hygiene (Cambridge) 38, 732±749.

Nakata, S. and Kimura, T. (1985) Effect of ingested toxic bean

lectins on the gastrointestinal tract in the rat. Journal of

Nutrition 115, 1621±1629.

Nakata, S. and Kimura, T. (1986) Behaviour of ingested concana-

valin A in the gastrointestinal tract of the rat. Agriculture

Biological Chemistry 50, 645±649.

Naughton, P.J., Grant, G., Ewen, S.W.B., Spencer, R.J., Brown,

D.S., Pusztai, A. and Bardocz, S. (1995) Salmonella typhimur-

ium and Salmonella enteritidis induce gut growth and increase

the polyamine content of the rat small intestine in vivo. FEMS

Immunology and Medical Microbiology 12, 251±258.

Naughton, P.J., Grant, G., Spencer, R.J., Bardocz, S. and

Pusztai, A. (1996) A rat model of infection by Salmonella typhi-

murium or Salmonella enteritidis. Journal of Applied Bacteriology

81, 651±656.

Ofek, I., Rest, R. and Sharon, N. (1992) Nonopsonic phagocytosis

of microorganisms; phagocytes use several molecular mechan-

isms to recognise, bind and eventually kill microorganisms.

American Society of Microbiology News 58, 429±435.

Ofek, I. and Sharon, N. (1988) Lectinophagocytosis; a molecular

mechanism of recognition between cell surface sugars and lec-

tins in the phagocytosis of bacteria. Infection and Immunity 56,

539±547.

Oliveira, J.T.A., vasconcelos, I.M., Gondim, M.J.L., Cavada,

B.S., Moreira, R.A., Santos, C.F. and Moreira, L.I.M. (1994)

Canavalia brasiliensis seeds. Protein quality and nutritional

implications of dietary lectin. Journal of the Science of Food and

Agriculture 64, 417±424.

Pistole, T.G. (1981) Interaction of Bacteria and Fungi with lectins

and lectin-like substances. Annual Review of Microbiology 35,

85±112.

Porter, J. and Pickup, R.N. (1998) Separation of natural popula-

tions of coliform bacteria from freshwater and sewage by mag-

net bead sorting. Journal of Microbiology Methods 33, 221±226.

Poschet, J.F. and Fairclough, P.D. (1999) Effect of lectins on

Salmonella typhimurium invasion in cultured human intestinal

cells. Gut 44 (S1), W252.

Pusztai, A., Ewen, S.W.B., Grant, G., Peumans, W.J., Van

Damme, E.J.M., Coates, M.E. and Bardocz, S. (1995) Lectins

and also bacteria modify the glycosylation of gut surface recep-

tors in the rat. Glycoconjugate Journal 12, 22±35.

Pusztai, A., Grant, G., King, T.P. and Clarke, E.M.W. (1990)

Chemical probiosis. In Recent Advances in Animal Nutrition ed.

Haresign, W. and Cole, D.J.A. pp. 47±60. London:

Butterworth Scienti®c Press.

Pusztai, A., Grant, G., Spencer, R.J., Duguid, T.J., Brown, D.S.,

Ewen, S.W.B., Peumans, W.J., Van Damme, E.J.M. and

Bardocz, S. (1993) Kidney bean lectin-induced Escherichia coli

overgrowth in the small intestine is blocked by GNA, a man-

nose speci®c lectin. Journal of Applied Bacteriology 75, 360±368.

Pusztai, A., Koninkx, J., Hendricks, H., Kok, W., Hulscher, S.,

Van Damme, E.J.M., Peumans, W.J., Grant, G. and Bardocz,

S. (1996) Effect of the insecticidal Galanthus nivalis agglutinin

on metabolism and the activities of brush border enzymes in

the rat small intestine. Journal of Nutritional Biochemistry 7,

677±682.

Rehmani, S.F. and Spradbrow, P.B. (1995) The contribution of

lectins to the interaction between oral Newcastle disease vaccine

and grains. Veterinary Microbiology 46, 55±62.

Sharon, N. and Lis, H. (1989) Lectins as cell recognition mole-

cules. Science 246, 227±234.

Sharon, N. (1993) Lectin-carbohydrate complexes of plants and

animals: an atomic view. Trends in Biochemical Science 18, 221±

226.

Sheriff, S., Chang, C.Y.Y. and Ezekowitz, R.A.B. (1994) The

structure of mouse mannose binding protein. Nature Structural

Biology 1, 789±794.

Sitkovsky, M.V., Pasternack, M.S. and Eisen, H.N. (1982)

Inhibition of cytotoxic T lymphocyte activity by concanavalin

A. Journal of Immunology 129, 1372±1376.

Summer, J.B. and Howell, S.F. (1936) The identi®cation of the

hemagglutinin of the jack bean with Concanavalin A. Journal of

Bacteriology 32, 227±237.

Thorns, C.J. (1995) Salmonella ®mbriae: novel antigens in the

detection and control of Salmonella infections. British Veterinary

Journal 151, 643±658.

Van Asten, A.J., Zwaagstra, K.A., Baay, M.F., Kusters, J.G.,

Huis in't Veld, J.H. and Van der Zeijsr, B.A. (1995)

Identi®cation of the domain which determines the g,m serotype

of the ¯agellin of Salmonella enteritidis. Journal of Bacteriology

177, 1610±1613.

Van Damme, E.J.M., Allen, A.K. and Peumans, W.J. (1987)

Isolation and characterisation of a lectin with exclusive speci®-

city towards mannose from snowdrop (Galanthus nivalis) bulbs.

FEBS Letters 215, 140±144.

Weaver, L.T. and Bailey, D.S. (1987) Effect of lectin concanava-

lin A on the neonatal guinea pig gastrointestinal mucosa in vivo.

Journal of Pediatric Gastroenterology and Nutrition 6, 445±453.

Zhou, Z.X., Deng, Z.P. and Ding, J.Y. (1995) Role of glycoconju-

gates in adherence of Salmonella pullorum to the intestinal

epithelium of chicks. British Poultry Science 36, 79±86.



Documents

Modulation of Salmonella infection by the lectins of Canavalia ensiformis (Con A) and Galanthus nivalis (GNA) in a rat model in vivo