Utilization of transferrin-bound iron byListeria monocytogenes

Abstract

It has been demonstrated that under iron-restricted conditions, Listeria monocytogenes can utilize iron-loaded transferrin (Tf) from a range of species as its sole source of iron for growth. Human transferrin conjugated to horseradish-peroxidase (HRP-Tf) bound directly to whole cells of L. monocytogenes. This binding was blocked by apotransferrin indicating that the receptor can bind transferrin in either the iron-bound or iron-free form. Transferrin-binding was not host specific because both bovine and equine transferrin inhibited the binding of HRP-conjugated human transferrin. SDS-PAGE and Western blotting of bacterial surface extracts revealed the presence of a transferrin-binding protein of approximately 126 kDa.

The 126 kDa iron-regulated protein of Listeria monocytogenes is not a transferrin binding protein

“…The hsterta strains examined failed to react with HRP-HTF (Fig la) but did react with blotmylated-HTF (Fig lb) The N menmgtttdts poslttve control reacted and the E colt negattve control failed to react in both cases (Fig 1) A similar pattern of reactlvJty with blotlnylated HTF (Fig 2a) but lack of reactivity wtth HRP-HTF (FJg 3a) for llsterla strains was also apparent by Hartford et al [7] Consequently, the reaction with blotmylated-HTF supported the findings of Hartford et al [7] of a c 126 kDa transferrm binding protein m L monocytogenes NCTC 7973, whereas the lack of reactivity with HRP-HTF contradicted such a view A possible explanation was that the apparent reaction with blotmylated-HTF was not due to lnteracuon with the transferrm molecule In order to test this possibdlty separated bacterial proteins were reacted with biotmylated human lactoferrln, and reaction was seen with a band identical to the one that reacted with blotmylated-HTF (Fig 2b) An explanation of these findings was forthcoming when we demonstrated that development of a membrane previously reacted with HRP-HTF (Fig 3a) with HRP-streptavldm alone resulted m vlsualisatlon of the same c 126 kDa protein (Fig 3b) A membrane reacted with HRP-HTF but not HRP-streptavidm remained negatwe for the hstena strains examined (not shown) demonstrating that the reactions seen m Fig 3b are not due to late development The evidence shows that this proposed transferrm binding protein, which also appeared to bind btotmylated lactoferrm, is m fact a naturally biotmylated protein that binds streptavldln The mablhty to purify a 126 kDa protein by the transferrin affinity procedure further supports the conclusion that this protein is not a transferrm-bmding protein Although these findings explain the Western blot results of Hartford and colleagues [7], which they interpreted as binding to transferrin, what Is more difficult to explain is their reported binding to HRP-HTF in whole cell dot-blots We could not demonstrate this m a repeat of their method- 4 and 6 respectwely) and m non-restricted BHI (lanes 3, 5, and 7 respectively) reacted with (a) horseradish peroxadase conjugated human transfernn and (b) the same membrane following reaction with horseradish peroxldase streptavldm Arrow head the 126 kDa L monocytogenes protein * the 78 kDa transferrm binding protein 2 of N memngmdts ology or by Western blot This is unhkely to be due to strata differences as the same NCTC 7973 stram was also examined In concluston we can find no evidence in support of the view that L monocytogenes binds dtrectly to human transferrin or that thts is mediated through a 126 kDa protein…”

“…The strams of L monocytogenes used were NCTC 7973 (serotype la), which was the strain used by Hartford et al [7], NCTC 11994 (serotype 4b) both from the National Collection of Type Cultures (Cohndale, UK), and IA486C (serotype 4b, a gift from Dr B Wren, St Bartholomews Hospital, London, UK) A laboratory faecal ~solate of Eschenchta cob 0157 H7 was used as a negative control and N menmgmdts strain SD (B15 P1 16) described previously [11] as a positwe control All strains were cultured on Columbia horse blood agar or grown in brain-heart-infusion broth (BHI, Merck) at 37°C aerobically m 5% CO 2 Iron hmltatlon for cultures of L monocytogenes and E colt was achieved as described by Hartford et al [7] by addition of ethylenediamme dl-o-hydroxyphenylacetate (EDDHA, Sigma), treated by the method of Rogers [12], to a final concentration of 100 tzM Cultures were grown at 37°C with shaking in 15 ml of culture medium to late log phase Cells were then harvested by centrlfugation at 1200 × g at 4°C for 30 mm, and the resultant pellets washed once with phosphate buffered sahne (PBS), re-pelleted, and re-sus-pended in 500/xl of 10 mM Trls-HCL (pH 7 4) to concentrate Protein est~mation of each bacterml culture was performed by the method of Bradford [13] using the Pierce protein estimation kit (Pierce) N menmgtttdzs was grown under ~ron-hmitatlon by previously described methods [11] Transferrm and lactoferrm bmdmg assays Three methods were used to detect transferrm binding proteins, whole cell dot lmmuno-blots, Western blots, and transferrm affimty purification For whole-cell dot lmmuno-blots 15 /zl (1 /zg/ml) of bacterial cells harvested following growth under iron-sufficient or iron-restricted conditions were spotted onto nitrocellulose membrane (Schlelcher and Schuell), the membrane blocked with 3% blocking buffer (consisting of 3% w/v skimmed milk and 0 1% Tween 20 m PBS), and probed with either 1 /zg/ml horseradish peroxadase conjugated human transferrm (HRP-HTF, Stratech Scientific Ltd), or 1 mg/ml biotin conjugated human transferrin (B-HTF), or 1 mg/ml biotin conjugated human lactoferrm (B-HL) B-HTF and B-HL were prepared according to the method described by Schryvers [11] Development was with 4-chloro-1-napthol (Sigma) for HRP-HTF reactions and HRP-streptavldm (50 ng/ml) (Jackson Immuno Research Lab Inc,) followed by 4-chloro-lnapthol for B-HTF and B-HL reactions These experiments were repeated using 3% (w/v) gelatin as the blocking agent For Western blots 50 /~g of bacterial protein were somcated at 12 × 103 m~crons for 5 mm (MSE somprep 150), dissolved in SDS and boiled and separated by sodium dodecyl sulphate 10% polyacrylamlde gel electrophoresls (SDS-PAGE) and transferred to nitrocellulose as described m detail elsewhere [11] The membranes were blocked with blockang buffer containing milk, then probed and developed as described above for whole-cell dot blots In addition the membrane which was reacted with HRP-HTF and developed was re-probed with HRP-streptavldm without prior incubation with a blotmylated compound An ldentxcal membrane to the one that was reprobed was incubated m parallel without re-prob-mg to confirm that any positive reaction seen following this procedure was not due simply to a late developing reaction of the HRP-HTF…”

“…Ltstena monocytogenes, a facultatlve Gram-pos~twe mtracellular pathogen, ~s the causatwe agent of llsterlos~s and is most commonly transmitted by food [1] Infection, which has a high mortahty rate even after specific chemotherapy [1], is particularly severe m the lmmunocompromlsed and newborn [2] Outcomes of refection can range from induction of abortxon, sept~-caem~a, through to menlngms Although there Is substantial genetic dwerslty amongst natural pop-ulat~ons of L monocytogenes, most mfect~on associated w~th epidemics of hsterlosls in separate parts of the world have revolved only two clones of L monocytogenes, both of serotype 4b [3,4] number of the molecular determmants of L monocytogenes pathogenesls are known [5], pnmartly those of the haemolysm (hstenolysm), leclthtnase and p60, a putatwe mvasm Although these vtrulence factors will undoubtedly play a role in ttssue damage and lntracellulartsatton, httle is known of factors that wtll permtt survtval m the host allowing expression of these virulence factors One of the limmng factors for bactertal growth m vwo is the avallabdlty of free iron Systems for acquiring iron have been identified for most pathogens studied to date [6] For Haemophtlus mfluenzae and Netssena rnenmgtttdts, both of which cause menmg~tts, this revolves the production of lactoferrm and transferrm binding protems [6] Recent work had shown that some strains of L monocytogenes also produce a protein that will bind human transferrm [7] The role of tron m promoting growth and wrulence of L monocytogenes has been known for many years [8] and three iron acquisition systems have been described These include the reduction of iron to the ferrous form by a soluble reductase [9], the direct binding of ferric citrate by a citrate inducible receptor [10], and most recently by the demonstration of a cell surface transferrm bmdmg protein [7] However, this most recent finding contradicts those of Cowart and Foster [9] who were unable to demonstrate a direct interaction with L monocytogenes of fluorescamme-labelled iron proteins, including human transferrln We were also unable in a previous unpubhshed study to detect direct binding of human transferrm to L monocytogenes by methods we have successfully used previously to show such reactions with N menmgtttdts [11] The differences in findings could be due to strain variation or differences in methodology We undertook the following study to resolve this issue…”

Identification of iron-regulated genes of Bifidobacterium breve UCC2003 as a basis for controlled gene expression

“…10 The acquisition of iron from host proteins has been studied in just a handful of Gram-positive microbes. Some species employ surface receptors for proteins, such as transferrin or lactoferrin, 11,12 while other Gram-positive bacteria, for example Bacillus subtilis, utilize siderophores to obtain ferric iron. 13 The use of haem or heamo proteins mediated through dedicated surface receptors has also been reported.…”