FEMS Microbiology Letters 77 (1991) 283-288 01991 Federation of European Microbiological Societies 0378-1097/91/$03.50

ADONIS 037810979100058H

FEMSLE 04277

283

Isolation and characterisation of Haemophilus influenzae type b mutants defective in transferrin-binding and iron assimilation

Julie H o l l a n d 1,2 Kevin J. T o w n e r 1 and Paul Wil l iams 2

1 Department of Microbiology and PHLS Laboratory, University Hospital, Nottingham, and 2 Department of Pharmaceutical Sciences, University of Nottingham, Nottingham, U.K.

Received 24 July 1990 Revision received 6 September 1990

Accepted 17 September 1990

: Key words: Haemophilus influenzae type b; Transferfin; Iron; Streptonigrin

1. SUMMARY

The quinone antibiotic streptonigrin was used to select mutants of Haemophilus influenzae type b defective in human transferrin binding. Com- pared with the parent wild-type strain (JKP1), mutant JKP5 was unable to bind transferrin whilst mutant JKP4 showed reduced binding. JKP5 ap- peared to lack an approximately 72 kDa transfer- fin-binding protein. Unlike JKP1, neither JKP4 nor JKP5 were able to acquire iron from human transferfin but their ability to use a variety of other iron and haem compounds as iron sources was unaffected. Such mutants should prove useful in further elucidating the mechanism of transfer- fin iron-acquisition and its contribution to the virulence of H. influenzae.

2. I N T R O D U C T I O N

Iron is an essential nutrient for bacterial growth and the ability of bacteria to acquire iron from an

Correspondence to: J. Holland, Department of Microbiology, University Hospital, Nottingham NG7 2UH, U.K.

infected host is known to contribute to virulence [1,2]. In serum, free iron is normally unavail- able to an infecting bacterium since it is tightly bound to transferfin, an iron-transporting gtyco- protein [1,2]. Many Gram-negative bacteria (e.g. Escherichia coli and Pseudomonas aeruginosa) overcome iron-restriction by synthesizing sidero- phores. These are high affinity, low molecular mass iron chelators which are able to remove transferrin-bound iron and transport it back to the cell via specific outer-membrane protein receptors [1,2]. Other pathogens, notably Neisseria gonor- rhoeae and N. meningitidis acquire iron via a siderophore-independent mechanism that involves a direct interaction between transferrin and the bacterial cell surface [3,4].

Haemophilus influenzae type b is an important human pathogen responsible for meningitis and other diseases, particularly in young children. We have recently reported that this organism ex- presses an inducible siderophore-independent iron-acquisition mechanism that is dependent upon the direct interaction between human trans- ferrin and an iron-regulated outer membrane pro- tein of about 72 kDa [5,6]. In this report we now describe the isolation and initial characterization

284

of mutants defective in this potentially important virulence determinant.

3. MATERIALS A N D M E T H O D S

3.1. Bacteria and growth conditions H. influenzae type b strain Eagan (JKP1) was

kindly provided by Prof. E.R. Moxon. Depart- ment of Paediatrics, John Radcliffe Hospital, Ox- ford. U.K. and routinely subcultured on chocolate blood agar plates. For experiments in liquid media, bacteria were grown with aeration at 37°C in brain heart infusion broth supplemented with 2.0 /~g/ml N A D and 0.5 /~g/ml protoporphyrin IX (BHIPPIX broth). Long-term storage of JKP1 and its mutant derivatives was at 70°C in BHIPPIX broth supplemented with 20% glycerol. I ron re- striction was achieved as described previously [5,6], by the addition of 170 /~M ethylenediamine di-o- hydroxy phenylacetic acid (EDDA) to BHIPPIX agar plates and 100 /~M E D D A to BHIPPIX broth.

3.2. Mutant selection procedures The antibiotic streptonigrin has been used to

isolate mutants defective in both siderophore-de- pendent [7] and siderophore-independent iron-up- take systems [8]. This quinone antibiotic reacts with intracellular iron pools to form hydroxyl radicals that cause D N A damage [9,10]. Mutants which are defective in high affinity iron assimila- tion will have low intracellular iron levels and should be relatively resistant to the action of streptonigrin. Two methods of mutant selection were used: (1) Direct selection. 0.1 ml portions of an over-

night BHIPPIX culture of JKP1 were plated on to chocolate blood agar plates containing 0.2 /~g/ml streptonigrin (Sigma). Colonies were counted after incubation for 24 h at 37°C. Total viable counts were similarly de- termined in the absence of streptonigrin.

(2) Enrichment. An overnight culture of JKP1 was subcultured into BHIPPIX broth contain- ing 0 .05/~g/ml streptonigrin. Following over-

night incubation with aeration at 37°C, 0.1 ml of this culture was subcultured into BHIPPIX broth containing 0 .2 /~g /ml streptonigrin. The cells were grown, with aeration, overnight at 37°C. 0.1 ml portions of this culture were then spread on to chocolate Mood agar plates con- tainmg 0 .2 /~g /ml streptonigrin. Colonies were counted after incubation for 24 h at 37°C. Total viable counts were determined as de- scribed above.

3.3. Assay for transferrin-binding ability The ability of H. influenzae to bind human

transferrin was evaluated using: (a) the solid-phase dot enzyme assay described previously [6] or (b) Western blotting of cell envelopes. These were enriched for transferrin binding activity as de- scribed by Schryvers [11] and electrophoresed on 10% sodium dodecyl sulphate (SDS)-polyacryl- amide gels using the modified conditions outlined in ref. 6. Dot blots were quantified using a Shimadzu CS-9000 scanning densitometer.

3.4. Ability of transferrin to restore growth m iron- restricted liquid media

Strains under test were inoculated into 250-ml flasks containing (i) 25 ml BHIPPIX broth only; (ii) 25 ml BHIPPIX broth plus 100 /~M EDDA (iron-restricted conditions); (iii) 25 ml BHIPPIX broth plus 100/~M E D D A plus 90% iron-saturated human transferrin (200 ~g/ml) . The flasks were incubated at 37°C and 220 rpm on a rotary shaker. Growth was monitored by measuring the optical density of the cultures at 600 nm.

3.5. Utilisation of iron sources The ability of H. influenzae to use 90% iron-

saturated human transferrin and a variety of other iron sources was examined with a plate bioassay [6,12]. I ron chelates and haem complexes were also prepared as described previously and used at concentrations of 20 m M or 1 m g / m l respectively [12]. Where required, human apotransferrin, al- bumin and haptoglobin were used as negative controls. Ferric nitrate (20 mM) was included as a positive control.

4. RESULTS

4.1. Isolation of streptonigrin-resistant mutants de- fective in transferrin-binding ability

Streptonigrin-resistant mutants derived from JKP1 were obtained by both of the selection pro- cedures described in SECTION 3.2. Mutants were obtained following direct selection at a frequency of 3 x 10 -7. A slightly enhanced frequency of 8 X 10 -7 w a s obtained following enrichment in increasing concentrations of streptonigrin. A total of 30 streptonigrin-resistant mutants were purified and maintained on streptonigrin-c0ntaining medi- um. Using the solid-phase dot enzyme assay, two mutants were found to be altered in transferrin- binding ability (Fig. 1). Following growth under iron-restricted conditions, mutant JKP5 was un- able to bind human transferrin whilst JKP4 ex- hibited reduced binding ability. Densitometric scanning experiments revealed that JKP4 bound 70% less transferrin than the parent strain JKP1.

4.2. Examination of mutants for transferrin-binding proteins

When compared with JKP1, no major alter- ations in the cell envelope protein profiles of JKP4 or JKP5 were apparent in Coomassie blue stained SDS-polyacrylamide gels (Fig. 2, lanes B-E). In

A B C + - - + - - + - -

N

Fig. 1. Solid-phase dot enzyme assay showing the binding of HRP-human transferrin conjugate to whole cells of H . in -

f l u e n z a e type b strain JKP1 (A) and mutants JKP4 (B) and JKP5 (C) immobilised on nitrocellulose. Bacteria were cultured in iron-sufficient conditions (+) in BHIPPIX or in iron-re-

stricted conditions ( - ) in BHIPPIX containing EDDA.

285

205-

116- 98-

66-

45-

29-

A B C O E Fig. 2. SOS-PAGE and Western blot analysis of the transfer- fin-binding proteins of H . i n f l u e n z a e type b. Cell envelopes prepared from bacteria grown in BHIPPIX broth (JKPI: lanes B and F) or in BHIPPIX broth with EDDA (JKPI: lanes C and G; JKP4: lanes D and H; JKP5: lanes E and I) were electrophoresed on 10% SDS-polyacrylamide gels and stained with Coomassie blue (lanes A-E) or transferred to nitrocellu- lose and probed with the HRP-transferrin conjugate (lanes F-I). Lane A shows the molecular mass marker proteins

(kDa).

Western blotting experiments using a horseradish peroxidase (HRP)-transferrin conjugate as a probe (Fig. 2, lanes F- I ) , JKP1 grown under iron-re- stricted conditions was found to possess a trans- ferrin-binding protein of approximately 72 kDa (Fig. 2, lane G). This protein has previously been reported to be a minor, iron-repressible outer membrane protein which is not easily visualized in Coomassie blue stained gels [6]. In JKP5 grown under iron-restricted conditions (lane I), or in JKP1 grown under iron-sufficient conditions (lane F), this protein was not observed. However, JKP4 grown under iron-restricted conditions (lane H) appeared to retain the 72 kDa protein, but at reduced levels.

4.3. Utilisation of transferrin-bound iron in liquid culture

The ability of human transferrin to restore the growth of JKP1, JKP4 and JKP5 was evaluated in liquid culture. The results in Fig. 3 show that addition of 90% iron-saturated transferrin re-

286

E e-" O O

0)

t -

O

<

A

0.6

0.5

0.4

0.3

0.2

0 2 4 6 Time (h)

E t -

O O LO

¢J ( , -

o

..Q ,<

B

0.6"

0.5"

0.4"

0.3"

0.2"

J

0 2 4 6

Time (h)

E 0.7" c- o 0.6' o ~0 0 .5" , * - - .

0.4- CD

t - 0 .3 -

-e 0.2 o

03 .o 0.1 ,<

0.0

C

i 1 i • i

0 2 4 6 0

Time (h)

Fig 3. Growth kinetics Of H. influenzae type b strain JKP1 (A) and mutants JKP4 (B) and JKP5 (C) in BHIPPIX broth. (O), control, no additions; (11), addition of 100 /~M EDDA; (©), addition of 100/~M EDDA and 200 g.g/ml 90% iron-saturated

human transferrin.

versed the growth inhibition imposed by E D D A on JKP1 (Fig. 3A) but not on mutants JKP4 (Fig. 3B) Or JKP5 (Fig. 3C).

4. 4. Utilisation of iron sources in plate bioassays The ability of JKP4 and JKP5 to use 90%

iron-saturated human transferrin, ferric nitrate, ferric citrate, ferric nitrilotriacetate, ferric 2,3 di- hydroxybenzoate, haem, haemoglobin, haemoglo- bin complexed with haptoglobin or haem bound to albumin as iron sources was evaluated using a plate bioassay. The parent strain JKP1 was capa- ble of using all of these compounds as iron sources. The only apparent difference between JKP1 and the mutants was an inability to acquire transfer- rin-bound iron.

5. DISCUSSION

The results presented in this paper demonstrate that streptonigrin selection is an effective means of isolating H. influenzae mutants defective in the binding of human serum transferrin. The loss of transferrin-binding ability in mutant JKP5 corre- lates with the inability of this mutant to use trans- ferrin as an iron source. The requirement for a direct surface interaction is in agreement with a previous study [6] in which it was shown that H. influenzae was unable to sequester iron from transferrin when the iron-binding protein was sep- arated from the bacterium by a dialysis mem- brane. In addition, the apparent loss of the 72 kDa transferrin-binding protein in JKP5 suggests that this protein probably functions as the recep- tor for human transferrin.

Several different types of mutants defective in transferrin-binding can be isolated by strep- tonigrin selection, as exemplified by the observed difference in binding ability between JKP4 and JKP5. However, although JKP4 retained 30% of the wild-type transferrin binding ability it was unable to use transferrin as its sole iron source. JKP4 may therefore also be defective in the steps following the cell surface binding of transferrin which lead to the assimilation of iron. The mecha- nics of this process have not yet been elucidated, but may involve reduction of bound ferric iron [13]. Alternatively, a conformational change in the structure of transferrin may occur on binding to the cell surface leading to release of iron. Studies are in progress to identify and clone the genes

involved in transferrin-iron acquisit ion in order to elucidate further details of this process.

Apar t f rom transferrin, H. infiuenzae is also able to use a variety of i ron salts, iron chelates and haem compounds as its sole source of iron [12]. The ability to acquire haem f rom haemoglobin as well as h a e m o g l o b i n / h a p t o g l o b i n and h a e m / a l - bumin complexes may be impor tant within mam- malian tissues as these compounds not only supply iron but also essential porphyrins [14]. Since streptonigrin may also select for mutat ions in a variety of other i ron-uptake systems, the ability of JKP4 and JKP5 to obtain iron f rom various iron and haem compounds was also examined. Inter- estingly, apart f rom an inability to obtain iron f rom transferrin, all of the compounds examined supplied iron to the mutants. Thus a l though H. influenzae possesses multiple i ron-uptake mecha- nisms, JKP4 and JKP5 appear to be defective only in the acquisition of iron f rom transferrin. The availability of mutants defective in transferrin-iron acquisition should enable the par t played by this system in the virulence of H. influenzae to be determined.

A C K N O W L E D G E M E N T

This work was supported by a grant f rom the Meningitis Trust, U.K.

R E F E R E N C E S

287

[11 Bullen, J.J. and Griffiths, E. (1987) Iron and Infection, Molecular, Physiological and Clinical Aspects, John Wiley and Sons, Chichester.

[2] Williams, P. (1988) Biochimie 70, 987-1011. [3] McKenna, W.R., Mickelsen, P.A., Sparling P.F. and Dyer,

D.W. (1988) Infect. Immun. 56, 785-791. [4] Tsai, J., Dyer, D.W. and Sparling P.F. (1988) Infect.

Immun. 56, 3132-3138. [5] Morton, D.J. and Williams, P. (1989) FEMS Microbiol.

Lett. 65, 123-128. [6] Morton, D.J. and Williams, P. (1990) J. Gen. Microbiol.

136, 927-933. [7] Braun, V., Gross, R., Koster, W. and Zimmermann, L.

(1983) Mol. Gen. Genet. 192, 131-139. [8] Dyer, D.W., McKenna, W., Woods, J.P. and Sparling,

P.F. (1987) Microb. Pathogen. 3, 351-363. [9] Yeowell, H.N. and White, J.R. (1982) Antimicrob. Agents

Chemother. 2, 961-968. [10] Cohen, M.S., Chai, Y., Britigan, B.E., McKenna, W.,

Adams, J., Svendsen, T., Bean, K., Hassett, D.J. and Sparling, P.F. (1987) Antimicrob. Agents. Chemother. 31, 1507-1513.

[11] Schryvers, A.B. (1989). J. Med. Microbiol. 29, 121-130. [12] Williams, P., Morton, D.J., Towner, K.J., Stevenson, P.

and Griffiths, E. (1990) J. Gen. Microbiol. in press. [13] Legrand, D., Mazurier, J., Montreuil and Spik, G. (1988)

Biochimie 70, 1185-1195. [14] Pidcock, K.A., Wooten, J.A., Daley, B.A. and Stull, T.L.

(1988) Infect. Immun. 56, 721-725.