Vol. 58, No. 1APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Jan. 1992, p. 42-470099-2240/92/010042-06$02.00/0

Characterization of Lactoferrin Binding by Aeromonas hydrophilaF. ASCENCIO,"12 A. LJUNGH,l AND T. WADSTROM1*

Department of Medical Microbiology, University of Lund, S-223 62 Lund, Sweden,1 andDepartment of Experimental Biology, Centro de Investigaciones Biologicas de

Baja California Sur, La Paz, Baja California Sur 23,000, Mexico2

Received 20 June 1991/Accepted 17 October 1991

Various lactoferrin preparations (iron-saturated and iron-depleted human milk lactoferrins and bovine milkand colostrum lactoferrins) were bound by Aeromonas hydrophila. Binding was (i) reversible (65% of boundlactoferrin was displaced by unlabeled lactoferrin), (ii) specific (lactoferrin but not other iron-containingglycoproteins such as ferritin, transferrin, hemoglobin, and myoglobin inhibited binding), and (iii) significantlyreduced by pepsin and neuraminidase treatment of the bacteria. The glycosidic domains of the lactoferrinmolecule seem to be involved in binding since precursor monosaccharides of the lactoferrin oligosaccharides(mannose, fucose, and galactose) and glycoproteins which have homologous glycosidic moieties similar to thoseof the lactoferrin oligosaccharides (asialofetuin or fetuin) strongly inhibited lactoferrin binding. A. hydrophilaalso binds transferrin, ferritin, cytochrome c, hemin, and Congo red. However, binding of these iron-containing compounds seems to involve bacterial surface components different from those required forlactoferrin binding. Expression of lactoferrin binding by A. hydrophila was influenced by culture conditions. Inaddition, there was an inverse relationship between lactoferrin binding and siderophore production by thebacterium.

The global distribution of Aeromonas species in manyaquatic ecosystems indicates the successful adaptation ofthe bacterium to such environments (14, 36). Also, Aeromo-nas hydrophila, Aeromonas sobria, and Aeromonas caviaeare opportunistic pathogens of poikilothermic (2, 17) andhomeothermic animals, including human beings (15). A.hydrophila is an important etiologic agent of gastrointestinalinfections in human beings (11, 35) and also can invade thebloodstream through defective intestinal mucosa (2). A.hydrophila produces virulence factors which enable thebacterium to colonize the host and to obtain nutrients andgrowth factors in vivo (1, 26, 28). To obtain iron required forgrowth, tissue-invasive Aeromonas species have developedspecific iron-sequestering mechanisms involving the produc-tion of extracellular hemolytic proteins (7, 19, 20, 25, 32, 33)and tryptophan- and phenylalanine-containing phenolate sid-erophores (6). In addition, Aeromonas salmonicida has beenreported (8) to utilize iron in iron-containing proteins such aslactoferrin and transferrin by mechanisms that require cellcontact.

Despite progress in defining the iron uptake mechanismsof Aeromonas species, there is a paucity of publishedinformation about the expression of Aeromonas cell surfacereceptors which can bind iron-containing proteins of the hostand which may be involved in the acquisition of iron by thebacterium. In this article, we report on the ability of A.hydrophila to bind lactoferrin, transferrin, ferritin, cy-tochrome c, and hemin.

MATERIALS AND METHODS

Bacterial strains and growth conditions. A total of 41 A.hydrophila strains were isolated from diseased fish andinfected humans (5). All media and chemicals were pur-chased from Difco (Detroit, Mich.), except blood agar baseno. 1 which was obtained from London Analytical & Bacte-

* Corresponding author.

riological Media Ltd., London. The bacteria usually weregrown at 32°C overnight on blood agar base supplementedwith 4% citrated horse blood, washed once with 0.02 Mpotassium phosphate buffer (pH 7.2) containing 0.15 M NaCl(PBS), and suspended in PBS to a density of 1010 cells perml. Some experiments were performed with bacteria grownin various culture media (indicated), at various temperatures(15, 25, 32, 37, and 42°C), in minimal medium broth adjustedto 0.15 M NaCl (MMB) (27) containing various concentra-tions of Ca2+ and in iron-depleted MMB containing variousconcentrations of the iron chelator 2',2-dipyridyl.

Binding and inhibition of `251-labeled lactoferrins and cop-per-containing proteins. All chemicals were purchased fromSigma Chemical Co. (St. Louis, Mo.). Lactoferrin frombovine and human milk (BLf and HLf, respectively) andfrom bovine colostrum, hemocyanin (HC), and hemolymph(HL) from Limulus polyphemus were labeled with 0.2 mCi of1251, by using Iodo-beads as previously described (21). Thebinding assay was performed as previously described forfibronectin binding by Escherichia coli (12). Briefly, 50 pAl of'25N-labeled protein (ca. 25,000 cpm) solution in PBS con-taining 0.1% bovine serum albumin was incubated with 100,ul of an A. hydrophila suspension (109 cells) in a polystyrenecentrifuge tube for 1 h at 20°C. Incubation mixtures notcontaining bacteria were used as background count controls.After adding 2 ml of ice-cold PBS containing 0.1% Tween 20,the mixtures were centrifuged (4,500 x g, 10 min, 4°C) andthe radioactivity of the pellets was measured in a gammacounter. Some binding assays were performed in the pres-ence of divalent cations (Ca2+, Mg2+, Mn2+, Fe2+) orEDTA (final concentration, 1 mM) and in the presence ofiron chelators [Desferal, ethylenediamine di-(o-hydroxyphe-nyl) acetic acid, and 2',2-dipyridyl] at a final concentration of1 mg/ml. The amount of 1251I-labeled protein bound to thebacteria was expressed as a percentage of the amount ofradiolabeled protein incubated with 109 bacteria.

Binding inhibition assays were performed by incubatingbacterial suspensions (100 pAl containing 109 cells) at 20°C for

42

LACTOFERRIN BINDING BY A. HYDROPHILA 43

1 h with 100 ,ug of unlabeled HLf and BLf with otheriron-binding or iron-containing proteins (ferritin [horsespleen], human transferrin, sheep hemoglobin, myoglobin[horse skeletal muscle], cytochrome c [horse heart], andbovine hemin) and with copper-containing proteins (HC andHL) before incubating the bacteria with 125I-lactoferrin. Thetreated cells were washed with 2 ml of PBS, suspended in100 ,u of buffer, and used in the binding assay describedabove. Inhibition values were calculated as the relativepercentage of 125I-labeled protein bound by bacteria sus-pended in PBS without inhibitor. Inhibition higher than 10%was considered significant.

Binding inhibition experiments were also performed withbacteria incubated with various monosaccharides (final con-centration, 100 mM) and glycoproteins (fetuin [fetal calfserum], human orosomucoid, bovine serum glycoprotein[fraction VI], and bovine submaxillary asialomucin; finalconcentration, 0.1%) before incubation with labeled lacto-ferrin.Enzymatic treatment of bacteria. All of the enzymes were

purchased from Sigma. A. hydrophila suspensions (100 [lIcontaining 109 cells) were incubated with 100 pug of pepsin,pronase E (from Streptomyces griseus), proteinase K (fromTritirachium album), trypsin (bovine pancreas), chymo-trypsin (bovine pancreas), and neuraminidase (from Vibriocholerae). The treatment conditions for each enzyme werethose described in the Worthington Enzyme Manual (37).The cells were washed three times with ice-cold PBS,suspended in the same buffer to the original cell concentra-tion, and used in the lactoferrin binding assay describedabove. Inhibition values were calculated as the relativepercentage of 125I-labeled protein bound to untreated bacte-rial cells suspended in PBS.

Assays for binding of ferritin, transferrin, cytochrome c,hemin, and Congo red. Bacteria grown on blood agar at 32°Cwere washed and suspended in PBS to a final concentrationof 1010 cells per ml. Samples (100 [LI) of suspension wereincubated, for various times at 37°C, with equal volumes ofsolutions of ferritin (1 mg/ml), transferrin (10 mg/ml), cy-tochrome c (1 mg/ml), hemin (100 p.g/ml), and Congo red(100 ,ug/ml), and the bacteria were sedimented by centrifu-gation (15,000 x g, 30 s). The supernatant fluids wereassayed spectrophotometrically for unbound proteins andhemin (A492) and for Congo red (A540). Binding was mani-fested as a decrease in absorbance relative to the absor-bances of the ferritin, transferrin, hemin, and Congo redsolutions which were not incubated with bacteria.Assay for siderophores in culture supernatant fluids. The

chrome azurol S substrate was prepared and the culturesupernatant fluids were assayed for siderophores by themethods of Schwyn and Neilands (30). Briefly, samples (0.5ml) of culture supernatant fluids were incubated at 22°C for1 h in the dark with the chrome azurol S substrate (0.5 ml),and siderophore activity was estimated by measuring theA630 of the mixtures. The positive control was the ironchelator Desferal, and uninoculated broths and PBS werenegative controls.

Analysis of data. Samples were tested in duplicate, andeach experiment was done at least three times. Backgroundresidual radioactivity from incubation mixtures without bac-teria was less than 5%. Binding values below 10% of the total125I-labeled glycoprotein added were considered to be neg-ative. Standard deviations were less than 6%.

100'

C,zEz

;-I

I-

INw

80

60

40

20

00

0 20 60 100 140 180

TIME (min)

FIG. 1. Displacement of cell-bound '25l-labeled lactoferrin fromA. hydrophila A186 by unlabeled lactoferrin.

RESULTS

Characterization of lactoferrin binding. Approximately40% of the 41 A. hydrophila strains examined bound 125I1labeled lactoferrin. A. hydrophila A186 (isolated from ahuman gastrointestinal infection), which binds connectivetissue proteins (5) and which exhibited a moderate capacityto bind lactoferrin, was selected for further characterizationof lactoferrin binding.The percentages of binding of iron-depleted HLf, iron-

saturated HLf, BLf, and bovine colostrum lactoferrin were35, 26, 31, and 38%, respectively (data not shown). Sixty-five percent of the 125I-BLf bound by A. hydrophila could bedisplaced by incubating the bacteria with 100 ,ug of unlabeledBLf for 1 h (Fig. 1). Incubation of the bacteria with unlabeledHLf, BLf, HC, and HL before exposure to labeled HLf andBLf inhibited binding (Table 1). However, other iron-con-taining glycoproteins (transferrin, ferritin, hemoglobin, myo-globin, and cytochrome c) did not inhibit binding of HLf andBLf.

Binding of 125I-labeled BLf and HLf was significantlyreduced by pepsin and neuraminidase treatment of thebacteria and, to a lesser extent by trypsin, pronase E, andproteinase K treatment (Table 2).To examine the importance of the iron complexed with the

lactoferrin for lactoferrin binding by the bacterium, bindingassays were performed in the presence of iron chelators. Theiron chelator 2,2-dipyridyl inhibited more than 50% of thebinding of lactoferrin by A. hydrophila (Table 2). In addition,the presence of Fe2+ or Ca2+, but not Mg2+, Mn2+, orEDTA, in the binding assay solutions (final concentration, 1mM) enhanced the binding of 125I-lactoferrin by 110 and71%, respectively (Table 3).The carbohydrate moieties of lactoferrin also seem to be

involved with binding, since an excess of the monosaccha-ride precursors of the oligosaccharide moiety of lactoferrin(i.e., fucose, galactose, and mannose), as well as asialofetuin

VOL. 58, 1992

44 ASCENCIO ET AL.

TABLE 1. Inhibition of binding of lactoferrin and copper-containing proteins by A. hydrophila A186

% Inhibition of binding of 25[-Inhibitor labeled:

BLf HLf HC HL

Iron-binding glycoproteinsBLf 62.2 65.2 66.2 70.0HLf 55.6 51.5 52.7 57.9Ferritin 0.0 0.0 0.0 1.6Transferrin 3.2 5.3 6.3 13.0

MetalloproteinsHemoglobin 5.6 0.0 52.2 46.0Myoglobin 0.0 11.0 0.0 8.7HL 3.0 11.0 60.5 56.5HC 28.0 18.0 67.4 64.4Cytochrome c 9.5 n.d. 0.0 10.6Hemin 56.0 64.0 34.3 34.2Ovalbumin 16.0 2.3 n.d. n.d.

CarbohydratesFucose 42.8 33.1 29.0 20.0Galactose 36.7 25.3 21.0 11.4N-acetyl-D-galactosamine 34.3 37.5 51.0 29.0Mannose 31.3 27.0 28.0 19.8

GlycoproteinsAsialofetuin 43.0 45.1 67.4 57.4Fetuin 26.0 23.1 52.7 46.6Submaxillary asialomucin 15.6 10.6 16.4 9.2Bovine serum glycoprotein 10.6 0.0 0.0 0.0Orosomucoid 2.0 8.3 0.0 0.0

or fetuin (which have glycan branches similar to lactoferrinoligosaccharides) (33) inhibit lactoferrin binding (Table 1). Inaddition, A. hydrophila binds the copper-containing glyco-proteins HC and HL which contain oligosaccharide moietiesanalogous to those of the lactoferrin (Table 1).

Effect of culture conditions on lactoferrin binding. Lacto-ferrin binding by A. hydrophila A186 was greatly influencedby growth conditions such as type of culture medium (Fig.2). Optimal binding of HLf, BLf, HC, and HL was observedat 15°C and decreased (31, 40, 41, and 44%, respectively) ina linear fashion from 15 to 42°C (data not shown). Inaddition, binding of HLf, BLf, HC, and HL was optimum at4.5 mM CaCl (data not shown). Bacteria grown in variousbroth media expressed an inverse relationship between125I-lactoferrin binding and siderophore activity. Bacteriagrown in media which enhanced expression of binding

TABLE 2. Effect of enzymes and iron chelators on binding oflactoferrin and copper-containing proteins by A. hydrophila A186

% Inhibition of binding of 251I-labeled:Treatment

BLf HLf HL

EnzymesChymotrypsin 6.6 14.9 28.8Pepsin 36.4 34.6 27.3Proteinase 16.0 15.8 40.5Proteinase K 12.7 19.0 9.5Trypsin 14.3 16.6 21.6Neuraminidase 36.4 33.0 24.2

Iron chelatorDesferal 26.6 13.0 14.0EDDHAa 14.4 6.4 3.52',2-Dipyridyl 49.0 57.2 38.5a EDDHA, ethylenediamine di-(o-hydroxyphenyl)acetic acid.

TABLE 3. Effect of divalent cations on the binding of125I-lactoferrin by A. hydrophila A186

% Binding of 125I-labeled:Addition to assay

HLf BLf

None (control) 35 31Ca2+ 60 53Fe2+ 74 67Mg2+ 37 33Mn2+ 34 31EDTA 33 30

activity exhibited depressed siderophore production (Table4).

Binding of ferritin, transferrin, cytochrome c, hemin, andCongo red by A. hydrophila. We have previously shown thatA. hydrophila binds to transferrin immobilized on latexbeads (3). However, the bacterium, which bound 1251-lacto-ferrin, did not commonly bind 125I-labeled transferrin, fer-ritin, cytochrome c, hemoglobin, or myoglobin (data not

100

80

c|r 600

40Cf 400

.0 20aT

CM

. 80

0co0 60

02o 40

201

0J

HLf BLf

HC HL

Growth mediaFIG. 2. Lactoferrin binding and binding of copper-containing

proteins by A. hydrophila A186 grown in various culture media.Abbreviations: BA, blood agar base supplemented with 4% citratedhorse blood; McC, MacConkey agar; TSA, tryptic soy agar; T-80,tryptic soy agar supplemented with 2% Tween 80; DHBA, trypticsoy agar supplemented with 5% defibrinated horse blood; CFA,colonization factor antigen agar (10); BHI, brain heart infusionbroth; TSB, tryptic soy broth; PW, peptone water; CFA-B, coloni-zation factor antigen broth (10); CDB, chemically defined broth forAeromonas spp. (27); CIB-4, broth for marine bacteria (4); MH,nutrient-rich broth for marine bacteria (4); M9/CA, minimal saltsplus Casamino Acids broth (1); 2216, 2216 marine broth mediumadjusted to 0.15 M NaCl (38).

APPL. ENVIRON. MICROBIOL.

LACTOFERRIN BINDING BY A. HYDROPHILA 45

60

0z

za

50

40

30

20

HLf0.6

04

0.5

0.4

0.3

0.2

0.1

0.0

0.~

A

A 0~~~~~

Qv

IIv

°b a

0- o

.__ . 0

1.60

1.55

.1.50

0

1.45

1.40

1.35

50 100 150 200 300

2,2- DIPYRIDYL(ug/ml)

FIG. 3. Binding of 125I-labeled lactoferrin and copper-containingproteins by bacteria grown in MMB depleted of iron by the additionof various concentrations of 2,2-dipyridyl.

shown). To rule out the possibility that the 125I-labelingprocedure deleteriously affected certain domains on theseproteins and eliminated their ability to be bound by thebacterium, we incubated A. hydrophila with unlabeled fer-ritin, transferrin, and cytochrome c, as well as with heminand Congo red, and measured the adsorption of thesecompounds. A. hydrophila A186 bound unlabeled transfer-rin, ferritin, cytochrome, hemin, and Congo red (Fig. 4).

DISCUSSION

A. hydrophila and related species (A. caviae and A.sobria) have at least two mechanisms for acquiring host iron.The first is a siderophore (amonabactin)-dependent systemfor extracting iron from Fe-transferrin (6), and the second isa siderophore-independent system for the utilization of hostheme compounds (i.e., hemoglobin-haptoglobin complexes)

TABLE 4. Siderophore activity and binding of lactoferrin andiron containing proteins by A. hydrophila A186

grown in different media

Culture fluid Siderophore % Binding of "25I-labeled:sulpernatanta actiVityb

(A630) BLf HLf HC HL

TSB 0.292 11.3 9.1 9.3 11.0BHI 0.323 11.5 10.1 9.7 13.2CDB 0.598 31.0 25.2 26.2 37.2M9/CA 0.632 40.8 32.5 26.8 35.3MMB 0.635 47.2 35.8 27.0 39.5

a TSB, tryptic soy broth; BHI, brain heart infusion broth; CDB, chemicallydefined broth for Aeromonas spp.; M9/CA, minimal salts plus Casanino Acidsbroth.

b Siderophore activity is manifested as a decrease in A630. The absorbancevalues of the uninoculated media (negative controls) were 0.60 to 0.65, and theabsorbance value for the positive control Desferal (100 p.g/ml) was 0.20. Noneof these uninoculated media interfered with the CAS assay. The experimentalprotocols are described in Materials and Methods.

0 10 20 30 40 50 60

TIME (min)FIG. 4. Binding of ferritin, transferrin, cytochrome c, hemin,

and Congo red by A. hydrophila A186. Binding conditions aredescribed in Materials and Methods. Symbols: 0, ferritin; A,transferrin; O, cytochrome c; A, hemin; 0, Congo red.

(22). Similar mechanisms for iron acquisition from transfer-rin and lactoferrin have been reported for A. salmonicida (8),where direct contact between bacterial cell surface receptorsand the iron-complexing protein is required. At the presenttime, it is not known whether one or both of the ironacquisition systems functions in vivo during Aeromonasinfections.The results presented here indicate that binding of lacto-

ferrin by A. hydrophila is specific and reversible. Thecarbohydrate moiety of the lactoferrin molecule is involvedin the binding process as revealed by inhibition assays withhomologous glycosides (Table 1). A. hydrophila A186 didnot show significant differences in its ability to bind various1251I-lactoferrin preparations isolated from different sources.Despite the fact that the complex oligosaccharide structuresof the lactoferrin molecule vary from one source to another(18, 23, 31), it seems that they are recognized by A. hydro-phila with the same binding affinity. Other iron-containingproteins, i.e., ferritin, transferrin, and cytochrome c, arealso bound by A. hydrophila (Fig. 4). However, the obser-vation that binding of BLf and HLf was not inhibited bythese iron-containing proteins (Table 1) suggests that bindingof the proteins involves different bacterial surface compo-nents than those required for lactoferrin binding. In addition,reduction of lactoferrin binding by pepsin and neuraminidasetreatment of the bacterium suggests that A. hydrophila cellsurface lactoferrin-binding components are proteins, andsince neuraminidase treatment affects lactoferrin binding,sialic acid and/or its negative charge may be present in thelactoferrin-binding components and are involved with bind-ing.Concerning the possible participation of iron, complexed

in the lactoferrin molecule, in the binding of lactoferrin by A.hydrophila, our results on (i) binding of iron-depleted andiron-saturated lactoferrin by the bacterium, (ii) binding oflactoferrin in the presence of iron chelators, and (iii) bindingof lactoferrin by the bacterium when there is an excess of

VOL. 58, 1992

w0---- - ----o HC

-I--,'

APPL. ENVIRON. MICROBIOL.

Fe2" ions in the incubation mixture are contradictory. Sincelactoferrin undergoes conformational changes, apparently ina redistribution of charges on the surface of the proteinmolecule when it has been saturated by iron (16), it seemsthat iron does not have any direct participation in thelactoferrin binding by A. hydrophila and that the findingsobserved are due to charge effects. However, iron is impor-tant for the expression of the lactoferrin-binding compo-nent(s) (Fig. 3).Deneer and Potter (9) have shown that Actinobacillus

(Haemophilus) pleuropneumoniae expresses two outermembrane proteins which are involved in iron acquisitionand which also bind hemin and Congo red (an aromatic dyestructurally similar to the heme of the hemin molecule). Onthe basis of these observations, we also evaluated the abilityof A. hydrophila cells to bind hemin and Congo red. Ourresults indicated that A. hydrophila binds these compounds(Fig. 4) and that hemin inhibits the binding of 125I-lactoferrinby A. hydrophila A186 (Table 1). A possible explanation forthis observation is that hemin, which is a highly hydrophobicmolecule (34), binds to A. hydrophila through hydrophobicinteractions (as is the case with Porphyromonas gingivalis)(13) and nonspecifically inhibits lactoferrin binding. Analternative explanation is that hemin competes with lacto-ferrin for the same binding site. Further studies to elucidatethe biochemical nature of the interaction between A. hydra-phila surface components and lactoferrin or hemin are nowin progress.Although the capacity of A. hydrophila to bind lactoferrin

has not been correlated with virulence, as is the case forNeisseria meningitidis (29), it seems plausible that the bac-terium might use such a source of host iron during in vivogrowth, as is suggested by its ability to grow in the low-Feenvironment found in vivo (7) and in vitro (Fig. 3). Furtherstudies are needed to determine whether lactoferrin-bindingcomponents are important in Aeromonas infections.We also studied the effect of culture conditions on the

capability of the bacterium to bind 1251I-lactoferrin, to eval-uate the possible influence of environmental physicochemi-cal factors on the ability of A. hydrophila to scavenge hostFe2+ from lactoferrin, and, in addition, the isolation, purifi-cation, and characterization of the lactoferrin-binding com-ponent(s) from the bacterium. Our results indicate that A.hydrophila grown in nutrient-poor media (MMB, minimalsalts plus Casamino Acids broth) at low temperatures (15°C)and in medium depleted of iron expresses the highest per-centage of binding of I251-lactoferrin (Fig. 2 and 3). On theother hand, bacteria grown in broth which enhances sidero-phore production do not bind I251-lactoferrin to the sameextent as do bacteria grown in broth which suppressessiderophore production (Table 4).

ACKNOWLEDGMENTS

This work was supported by grant 16X04723 from the SwedishMedical Research Council and by grants from the Swedish Instituteand from the Swedish Board for Technical Development.We thank A. Kreger for his valuable comments and discussions.

REFERENCES1. Allan, B. J., and R. M. W. Stevenson. 1981. Extracellular

virulence factors of Aeromonas hydrophila in fish infections.Can. J. Microbiol. 27:1114-1122.

2. Altwegg, M., and H. K. Geiss. 1989. Aeromonas as a humanpathogen. Crit. Rev. Microbiol. 16:253-286.

3. Ascencio, F., P. Aleljung, and T. Wadstrom. 1990. Particleagglutination assay to identify fibronectin and collagen cell

surface receptors and lectins in Aeromonas and Vibrio species.Appl. Environ. Microbiol. 56:1926-1931.

4. Ascencio, F., A. L. Cortez, and J. L. Ochoa. 1988. Differences incell surface hydrophobicity of pigmented halobacteria and cor-responding achromatic strains evidenced by adsorption ontophenyl-Sepharose. Microbios Lett. 37:131-135.

5. Ascencio, F., A. Ljungh, and T. Wadstromn. 1991. Comparativestudy of extracellular matrix protein binding to Aeromonashydrophila isolated from diseased fish and human infection.Microbios 65:135-146.

6. Barghouthi, S., R. Young, M. 0. J. Olson, J. E. L. Arceneaux,L. W. Clem, and B. R. Byres. 1989. Amonabactin, a noveltryptophan- or phenylalanine-containing phenolate siderophorein Aeromonas hydrophila. J. Bacteriol. 171:1811-1816.

7. Buckley, J. T., L. N. Halasa, K. D. Lund, and S. McIntyre. 1981.Purification and some properties of the hemolytic toxin aerol-ysin. Can. J. Biochem. 59:430-435.

8. Chart, H., and T. J. Trust. 1983. Acquisition of iron fromAeromonas salmonicida. J. Bacteriol. 156:758-764.

9. Deneer, H. G., and A. A. Potter. 1989. Effect of iron restrictionon the outer membrane proteins of Actinobacillus (Haemophi-lus) pleuropneumoniae. Infect. Immun. 57:798-804.

10. Evans, D. G., D. J. Evans, Jr., and W. Tjoa. 1977. Haemagglu-tination of human group A erythrocytes by enterotoxigenicEscherichia coli isolates from adults with diarrhea: correlationwith colonization factor. Infect. Immun. -18:330-337.

11. Figueroa, G., H. Galeno, V. Soto, M. Troncoso, V. Hinrichsen,and A. Yudelevich. 1988. Enteropathogenicity of Aeromonasspecies isolated from infants: a cohort study. J. Infection17:205-213.

12. Froman, G., L. M. Switalski, A. Faris, T. Wadstrom, and M.Hook. 1984. Binding of Escherichia coli to fibronectin. J. Biol.Chem. 23:14899-14905.

13. Grenier, D. 1991. Hemin-binding property of Porphyromonasgingivalis outer membranes. FEMS Microbiol. Lett. 77:45-50.

14. Hazen, T. C., C. B. Fliermans, R. P. Hirsh, and G. W. Esch.1978. Prevalence and distribution of Aeromonas hydrophila inthe United States. Appl. Environ. Microbiol. 36:731-738.

15. Khardori, N., and V. Fainstein. 1988. Aeromonas and Plesiomo-nas as etiological agents. Annu. Rev. Microbiol. 42:395-415.

16. Lane, R. S. 1971. DEAE-cellulose chromatography of humantransferrin: the effect of increasing iron saturation and copper(II) binding. Biochim. Biophys. Acta 243:193-202.

17. Larsen, J. L., and N. J. Jensen. 1977. An Aeromonas speciesimplicated in ulcer-disease of the cod (Gadus morhua). NordVet. Med. 29:199-211.

18. Leclerq, Y., G. Swatzki, J. M. Wieruszeski, J. Montreuil, and G.Spik. 1987. Primary structure of glycans from mouse serum milktransferrins. Biochem. J. 247:571-578.

19. Ljungh, A., and T. Wadstrom. 1982. Aeromonas toxins. Phar-macol. Ther. 15:339-354.

20. Ljungh, A., and T. Wadstrom. 1983. Toxins of Vibrio parahae-molyticus and Aeromonas hydrophila. J. Toxicol. Toxin Rev.1:257-307.

21. Markwell, M. A. K. 1982. A new solid-state reagent to iodinateproteins. I. Conditions for the efficient labeling of antiserum.Anal. Biochem. 125:427-432.

22. Massad, G., J. E. L. Arceneaux, and B. R. Byers. 1991.Acquisition of iron from host sources by mesophilic Aeromonasspecies. J. Gen. Microbiol. 137:237-241.

23. Matsumoto, A., H. Yoshima, S. Takasaki, and A. Kobata. 1982.Structural study of the sugar chains of human lactoferrin:finding of four novel complex-type asparagine-linked sugarchains. J. Biochem. 91:143-155.

24. Mevarech, M., and R. Werczberg. 1985. Genetic transfer inHalobacterium volcani. J. Bacteriol. 162:461-462.

25. Nomura, S., Fujino, M. Yamakawa, and E. Kawahara. 1988.Purification and characterization of salmolysin, an extracellularhemolytic toxin from Aeromonas salmonicida. J. Bacteriol.170:3694-3702.

26. Paniagua, C., 0. Rivero, J. Anguita, and G. Naharro. 1990.Pathogenicity factors and virulence for rainbow trout (Salmogairdneri) of motile Aeromonas spp. isolated from a river. J.

46 ASCENCIO ET AL.

LACTOFERRIN BINDING BY A. HYDROPHILA 47

Clin. Microbiol. 28:350-355.27. Pazzaglia, G., and R. B. Sack. A new chemically defined

medium for culturing of motile Aeromonas species, abstr.C-170, p. 351. Abstr. 87th Annu. Meet. Am. Soc. Microbiol.1987. American Society for Microbiology, Washington, D.C.

28. Santos, Y., A. E. Toranzo, J. L. Barja, T. F. Nieto, and T. G.Villa. 1988. Virulence properties and enterotoxin production ofAeromonas strains isolated from fish. Infect. Immun. 56:3285-3293.

29. Schryvers, A. B., and G. C. Gonzalez. 1989. Comparison of theability of different protein sources of iron to enhance Neisseriameningitidis infection in mice. Infect. Immun. 57:2425-2429.

30. Schwyn, B., and J. B. Neilands. 1987. Universal chemical assayfor detection of siderophores. Anal. Biochem. 160:47-56.

31. Spik, G., G. Strecker, B. Fournet, S. Bouquelet, and J. Mon-treuil. 1982. Primary structure of the glycan from lactoferrin.Eur. J. Biochem. 121:413-419.

32. Titball, R. W., and C. B. Munn. 1983. Partial purification andproperties of a haemolytic activity (T-lysin) from Aeromonas

salmonicida. FEMS Microbiol. Lett. 20:207-210.33. Titball, R., and C. B. Munn. 1985. The purification and some

properties of H-lysin from Aeromonas salmonicida. J. Gen.Microbiol. 131:1603-1609.

34. Tsutsui, K., and G. C. Mueller. 1982. Affinity chromatographyof hemin-binding proteins: an improved method for the synthe-sis of hemin-agarose. Anal. Biochem. 121:244-250.

35. Wadstrom, T., and A. Ljungh. 1988. Correlation between toxinformation and diarrhoea in patients infected with Aeromonas. J.Diarrhoeal Dis. Res. 6:113-119.

36. Williams, L. A., and P. A. La Rock. 1985. Temporal occurrenceof Vibrio species and Aeromonas hydrophila in estuarine sedi-ments. Appl. Environ. Microbiol. 50:1490-1495.

37. Worthington Biochemical Corp. 1978. Worthington, enzyme andrelated biochemicals. Worthington Biochemical Corp., Free-hold, N.J.

38. ZoBell, C. E. 1941. Studies on marine bacteria. I. The culturerequirements of heterotrophic aerobes. J. Mar. Res. 4:42-75.

VOL. 58, 1992