.n‘ 8}" fax}: “I?! 3% ,5 V1. ,‘.,;.f-lfI/‘} . any“ POIINL,‘ «‘2’? d". 415, ! pf if‘? 1.45; 23,125.51; 1'; "‘ J23.“ 1:3:99‘ 1;} It I“? If} :., “.3 “I {I}, {fit}? 'JLI I:} a“ 4"” u ) L‘:“{$" .jé‘ay :1? | ‘h :f'? :1 ‘ ’- l -n‘-‘ ' n" u‘ 5 1.1.3.2"! V '13}: 35.551, ({Ic‘V {v I 4. “..~ '15:;._"'}"..:. N" 3", .$:::‘:{“ w‘~.a‘o‘4 o (x 5’5? a." 5 a - flsfi.“ T" £311.; . J' 'N. . 5.5‘ v‘;};‘ . "'5???" .w .3 ffifi: . ‘ "P ‘ 623‘": ‘1" $55. }' m“ x h! . v‘n’VW .wuw’(~‘q u ,ij: i" 1‘ ‘ZI 1.)th 1'31. “A“ ”4‘l§ .I . w W4 ‘ .‘yflmw. tw’ .érw' {’W' n) 3" {”6" $35.13: }\ I I. w J." .\ ‘n‘ '3‘?” A}? 4.1.9 ‘.' w EV w ‘z‘ ‘ I . J ' ""353: ‘I 1:? .1. n“. VI I l‘\4:'\~ "“"’“ :‘:‘|;:‘ 5‘. q-‘nv . ~- ‘4 u . u?" 5 u' 1.“ 33.11. a‘n ' N. , ‘ 1:33] 155?: E I. " .;,gr.v mg.“ '1’" Y 2 p." ." 'o ’ t}'\~ ”I, 233:; 'II' gift“ Vt; :9? :NE: ".;-‘ q‘n'ilk’ ‘ ' ': 3'- 05”“! "V _'~‘/‘.11’M':-;W I I v '.~ lift, 1 f" 1, (€"('.'« ‘I‘l'v‘ gm; ’5' ’v ‘5‘. 5- ' ' u .u‘."‘. '1‘ (I “It-("43$ ‘ 5 "5;? $\O}A\ I “I ¢. { "QI (VG‘I‘; I I , " :45 13% L35 :1?! -\ u‘rnuu " “qr ”3",". -“"' WIV:‘:1‘::I“: mm}; 1.59"?” -:I .31"? 'thl {Flu/a -'4'4 ll 'Ifztr‘fi 4"ny“ {’15 ”mg 22.3,:512 .‘;J":.1'V.I:‘£H "V'I WV}; ." ' 5x :4‘15’5‘54 :75??? W?” ’55:"? I ' ' . ~3VT-‘3x‘?!’ (£351? . a! "I016? 15%)}: xii { (“:4 [3" . M Y fzg4£4 ' llv'l’fif‘J/u’ k ‘1'. s .‘l.l XV? fl-‘i‘fi' . . I - ov‘:( . ‘n‘v\"' V “’62:."- ‘1. J. ‘ ‘y ‘|( "EN -‘)' 17' ‘ u' u,(( .. _ W55-‘w;‘ .M f 4.72% ”94"? ngwu‘z‘ K v'l“ll \ I.‘ ”Iv/.1 1(- ' . 2,-‘5- ‘r' k: i 5. ' 3:? in“? l I 55.9 '4‘, ,4?! I";‘_’J.,: 1-. a .-,J . n‘h' c o -" u I“ .v‘ I I flic‘ I ‘1 '11.: 9":3 I . . 'I’l. M1194 y ~1- I -ul‘ llllllHllllllllllllllllllllllllllllllllllllllllllllllllill L 3 1293 00559 8168 LIBRARY Michigan State University This is to certify that the dissertation entitled VIRULENCE FACTORS ASSOCIATED WITH IRON ACQUISIT- ION AND PESTICIN RESISTANCE IN YERSINIAE presented by DANIEL J . SIKKEMA has been accepted towards fulfillment of the requirements for Ph. D. degree inMiCLthology WW Major professor I Date 12-16-88 MS U is an Affirmative Action/Eq ual Opportunity Institution 0-12771 IV1f31.J RETURNING MATERIALS: Place in book drop to Llaamuas remove this checkout from Ail-KSIIIL. your record. FINES will be charged if book is returned after the date stamped below. I CC TB 92 it} 9.3 ‘0575 VIRULENCE FACTORS ASSOCIATED WITH IRON ACQUISITION AND PESTICIN RESISTANCE IN YERSINIAE Daniel J. Sikkema A DISSERTATION Submitted to Michigan State university in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1988 VJ K") 5&73 ABSTRACT VIRULENCE FACTORS ASSOCIATED WITH IRON ACQUISITION AND PESTICIN RESISTANCE IN YERSINIAE By Daniel J. Sikkema The independent abilities of cells of Yersinia pestis, the causative agent of bubonic plague, to absorb exogenous pigments including hemin and Congo red (Pgm+) and to produce the bacteriocin pesticin with genetically linked invasive enzymes (Pst+) are estab- lished virulence factors of the species. Loss of the 10 kilobase pesticin plasmid (Pst-) while retaining pigmentation (Pgm+) results in strains that are sensitive to pesticin (Psts). Mutation to pesticin resistance (Pstr) results in concomitant conversion to Pgm-. Wild type but phenotypically non-pigmented Yersinia pseudotuberculosis and Yersinia enterocolitica may be Psts; nutation of the latter to P‘str also results in avirulence, unless manmalian serum iron transport pro- teins are saturated by injected iron. It is shown in this thesis that cells of Pgm+ X. pgstis strain KIM possess a distinct growth advantage over P911:- cells in a highly iron-deficient environment at 37°C. In + . . . accordance, Pgm cells synthesme a unique set of at least nine chromosomally—encoded iron-repressible outer-membrane peptides (Irps), one of which is the pesticin receptor. At least one additional Irp is mediated by the pesticin plasmid. Mutants isogenic for the pesticin receptor, the major Irp, can be obtained while retaining pigmentation (Pgmi, Pstr) during selection on Congo red-pesticin agar. The mutation 0 mutations per bacterium frequency for this reaction is about 5 x 10.1 per generation. Pgmf, Pstr cells are unable to replicate in highly iron-deficient environments at 37°C and thus resemble Pgm' cells. Resistance to pesticin in the enteropathogenic yersiniae does not pmomote a similar requirement for iron in gi§£9_but rather prevents these avirulent organisms from penetrating HeLa cells. Expression of Irps in these strains greatly enhances HeLa cell invasion. Growth of all yersiniae was enhanced when conalbumin agar was supplemented with hemopexin, hemoglobin, hemin, myoglobin, ferritin or Fe3+, but no stimulation was obtained with 23%, 50%” 75%.or 95% iron-saturated human transferrin or lactoferrin. It is concluded that synthesis of Irps in enteropathogenic yersiniae permits survival within iron-deficient mammalian extracellular environments and may promote uptake into the iron—enriched intracellular environment. ACMOFEEIBEMENTS I would like to thank my graduate conmittee: Dr. Neal Band, Dr. wendy Champness, Dr. Robert Hausinger, and Dr. James Jensen, for their advice and suggestions. Additionally, Dr. Elizabeth werner and Dr. Ronald Patterson were very helpful in their supply of HeLa cells. I would especially like to thank Dr. Robert R. Brubaker for his encourage- ment, guidance and support over the last four years. iv TABLE OF CONTENTS LI ST 0F TAKES O O O O O O O O 0 O O O O 0 LIST OF FIGURES . . . . . . . . . . . . . I mowml w 0 O O O O O O O O O O O O O 0 Literature Review . . . . . . . . . . Iron Privation . . Purine Biosynthesis . . . . . . Fraction I . . . . . . . . . . . Lcr Plasmid . . . . . . . . . . Pesticin Plasmid . . . . . . . . Pigmentation . . . . . . . . . . CHAPTER I ARTICLE: Resistance to pesticin, storage of iron, invasion of HELa cells by yersiniae . . . . Abstract............ Introduction . . . . . . . . . . Materials and Methods . . . . . Results . . . . . . . . . . . . Discussion . . . . . . . . . . . References . . . . . . . . . . . CHAPTER II ARTICLE: Outer-membrane peptides of yersiniae and EsCherichia coli associated with iron and sensitivity to pesticin Abstract . . . . . . . . . . . . Introduction . . . . . . . . . . Materials and Methods . . . . . Results . . . . . . . . . . . . Discussion . . . . . . . . . . . References . . . . . . . . . . . acquisition Page ix 20 21 21 21 22 24 26 28 29 31 34 38 54 60 Page CHAPTER III Utilization of biological iron compounds by yersiniae . . . 65 Introduction . . . . . . . . . . . . . . . . . . . . . 66 Materials and Methods . . . . . . . . . . . . . . . . 67 Results . . . . . . . . . . . . . . . . . . . . . . . 71 Iron Storage . . . . . . . . . . . . . . . . . . . . . 74 Siderophore Production . . . . . . . . . . . . . . . . 78 CHAPTER IV HeLa cell invasion by yersiniae is an iron-stress function 0 O I O O O O O O O O O O O O O O O O O O O O O 0 80 Introduction . . . . . . . . . . . . . . . . . . . . . 81 Materials and Methods . . . . . . . . . . . . . . . . 82 Results . . . . . . . . . . . . . . . . . . . . . . . 84 Discussion . . . . . . . . . . . . . . . . . . . . . . 87 WY 0 O O O O O O O O O O O O O O O O O O I O O O O O O O O 89 APPENDIX Isolation of a pesticin resistant, pigmented (Pgmf, Pstr) strain of Yersinia pestis . . . . . . . . . . . . . . . . . 91 LIST OF REFERENCES . . . . . . . . . . . . . . . . . . . . . . . 92 ‘vi Table LIST OF TABLES INTRODUCTION The LD50 of wildtype‘ and Pstr yersiniae for mice by various routes of infections ........ . . . Plasmidsinyersiniae. . . Accepted nomenclature and approximte molecular weights of Yops expressed by the yersiniae . . Virulence factors of Yersinia pestis and their effect on the LD50 for mice via various routes 0f infxtion O O O O O O O O O O O O O O O 0 CHAPTER I Properties of yersiniae strains . . . . . . . . . . Doubling times and maximal optical density of Pgm+ and Pgm" strains of X_. Estis at 37°C in chemically defined iron—deficient medium . . . . . . . . . . . Infectivity for HeLa cells of Pgm+ and Pgm- isolates of 1. Estis, PstS and Pstr isolates of _l_{_._ pseudotuberculosis and Psts and Pstr isolates ofLenterocolitica . . . . . . . . . . . CHAPTER II Properties of bacterial strains ...... . . . . Iron-repressible outer-membrane peptides . . . . . Iron—inducible outer-membrane peptides . . . . . Page 13 19 22 24 25 35 48 48 Table CHAPTER III Properties of bacterial strains . . . . . . Composition of conalbumin agar . . . . . . Composition of siderophore detection agar . CHAPTER IV PrOperties of bacterial strains tested for infectivity of HeLa cells . . . . . . . . . Infectivity for HeLa cells by yersiniae grown in the presence and absence of iron . Page 68 69 7O 82 85 Figure LISP OF FIGURES CHAPI‘ERI Growth comparison of Pgm+ and Pgm' cells of _‘x_'. Estis KUMA in chemically defined iron-extracted medium with and without supplemental Fe013 Growth comparison of Pgm+ and Pgm' cells of I. Estis KUMA in chemically defined iron-extracted medium with and without supplemental hemin . . . . . . . . . . . The effect of temperature upon the growth of Pgm’ nutants of 1. Estis in highly iron-deficient Outgrowth of Pgm+ cells from an artificial mixture of of 0.05% Pgm“ cells and 99.95% Pgm‘ untants in an iron-deficientmedium................ Comparative morphology of Pgm+ and Pgm' cells of l. Estis KUMA after a single transfer in iron- defiCient mdim O O O O O O O O O O O O O O O O 0 Comparison of growths between Pgm+ and Pgm- cells of y_. Estis KUMA, Psts and Pstr cells of _Y_. pseudotuberculosis FBI, and Psts and Pstr cells of l. enterocolitica WA during a second transfer in iron-deficientnedium................ CHAPTERII Electropherograns (two-dimensional) of outer membranes of Escherichia coli strain 93 grown at 37°C in the presence and absence of iron . . . . . . . . . . Electropherograms of outer nembranes of (A) f_ep, (B) Pstr, (C) FhuA, and (D) gig nutants of Escherichia coli strain 95 grown in iron-deficient ix Page 23 23 23 23 24 24 39 4O Figure 10. Growth of (A) wild type, (B) Pstr, (C) Leg, and (D) tonB cells of Escherichia coli strain 9! in iron-deficient medium or the same medium containing addedheminorexcessFeClB. . . . . . . . . . . Electropherograms of outer membranes of Yersinia enterocolitica strain WA grown (A) with excess FeCl , (B) in iron-deficient medium, and (C) guter membranes of an isogenic Pstr mutant grown in iron-deficient medium . . . . . . . . . . Electropherograms of outer membranes of Yersinia pseudotuberculosis strain WA grown (A) with excess FeCl3, (B) in iron-deficient medium, and (C) outer membranes of an isogenic Pstr nutant grown in iron-deficient medimn . . . . . . . . . Electropherograrrs of outer membranes of Pgm+, Pst+ cells of Yersinia pestis strain KIM grown at 37°C in the presence and absence of iron . . . . . Electropherograms of outer membranes of Pgm', Pst“ cells of Yersinia pestis strain KIM grown at 37°C in the presence and absence of iron . . . . . Electropherograms of outer membranes of Pgm+, Pst ', and Pgm‘, Pst- cells of _Y_. Estis strain KIM showing iron-repressible peptides independent of pesticinogeny . . . . . . . . . . . . . . . . . . Electropherograns of outer membranes of Pgm+, Pstr cells of X. pestis strain KIM lacking structures associated with absorption of pesticin . . . . . . Growth comparison of Pgm+ (Pst+ and Pst”), Pgm- (Pst:+ and Pst') and Pgm+, Pst", Pstr cells of X. Estis strain KIM in chemically defined CHAPTER III Growth of yersiniae on conalbumin agar supplemented with biological iron-containing compounds . . . . . Separation of 55Fe-containing compounds from the bulk cytoplasmic protein obtained from Pgm+ cells Of X. Estis KIM O O O I O O O O O O O O O O O O O Page 41 44 46 47 50 51 52 53 73 75 Page SDS-PAGE of partially purified cytoplasmic iron storage compound . . . . . . . . . . . . . . . . . . 76 Growth of yersiniae and Escherichia coli on siderOphore detection agar . . . . . . . . . . . . . 79 CHAPTER IV Time course of infectivity for HeLa cells by X, pseudotuberculosis PBl previously rown at 37°C in the presence or absence of Fe + . . . . . . . 85 xi LITERATURE REVIEW The genus Yersinia, a member of the enterobacteriaceae, includes 1, pseudotuberculosis and X, enterocolitica, important causes of human diarrheal diseases, Z, (Pasteurella) pestis, the causative agent of bubonic plague, and others. Plague is an infection of wild rodents, transmitted from one rodent to another and occasionally from rodents to humans by the bites of fleas. An estimated 100-150 million human fatalities occurred in Europe between the Plague of Justinian (”540- 590 A.D.) and the second great pandemic, the black death (1347- 1350 A.D.); a total greater than that occurring in all of the wars of the world combined (19,36). Plague transmission is most frequently the result of a flea obtaining a blood meal from an infected rodent. Ingested cells of X, pestis multiply in the gut of the flea and, helped by the bacterial enzyme coagulase, block the flea proventriculus so that no food can pass through. Subsequently, the "blocked," hungry flea bites its next victim, and the aspirated blood, contaminated with X, pestis from.the flea, is regurgitated into the bite wound. The bacteria rapidly dis- seminate from the site of the flea bite by entering the dermal lym- phatic system and become localized in lymph nodes. An intense hemor— rhagic inflammation develops in the enlarged lymph nodes, whiCh may undergo necrosis and cause spillover into the circulatory and 2 reticuloendothelial system. Systemic disease is rapidly fatal unless antibiotics are administered innediately. Another manifestation of plague is the pulmonary or pneumonic form. Primary pneumonic plague results from inhalation of droplet nuclei from a coughing patient but may also arise secondarily due to localization, within the lungs, of organisms from the lymphatic and/or circulatory systems. Mortality rates of nearly 100% are reported for pneumonic plague unless prompt antibiotic therapy is initiated. Enteropathogenic yersiniae are of considerably reduced virulence when compared to 1. Estis (see Table 1). Transmission of X. gseudotuberculosis and 1. enterocolitica is primarily via the oral/ fecal route, thereby negating the need for the highly invasive prOper- ties exhibited by X. pestis. An innoculum of 108-109 yersiniae must enter the alimentary tract to cause human disease (81). The yersiniae multiply in the gastrointestinal mucosa and may invade into the sur- rounding lymphatic regions. Inflanmation, ulceration, and diarrhea with leukocytes in the feces are the usual event; in most cases, it is self limited. Severe cases may lead to bacteremia or meningitis but this is a rare event usually limited to inmunocompromised individuals. The yersiniae model has been extremely important to medical microbiologists, allowing the resolution of many mechanisms of bacte— rial virulence at the molecular level (19—22,24,30,106). Success of the model centers around the close taxonomic relationship shared between the yersiniae and Escherichia coli. Additionally, the high virulence of yersiniae for mice allows pathogenicity to be measured in simple terns of lethality rather than cumbersome determinations of in vivo proliferation . 3 TABLE l. The ID 0 of wildtype yersiniae for mice by various routes of infect ons.a Route of infectionb Yersinia species s.c. i.p. i.v. l. Estis <101 <101 <101 1. pseudotuberculosis 1.6 x 105 2.4 x 104 2.0 x 101 l. enterocolitica 2.3 x 103 2.1 x 102 1.0 x 102 a. Data taken from Une and Brubaker (132). b. Abbreviations are as follows: s.c., subcutaneous injection i.p., intraperitoneal injection i.v., intravenous injection 4 Numerous advances first made with the yersiniae model are directly associated with the conclusions presented in this thesis, including demonstrations that ( i) plasmids can mediate bacterial virulence (26,58,83) (see Table 2 ); (ii) in y_i\_r_o growth of pathogenic organisms is limited by availability of iron (see Table 2) (41,77, 78,79); (iii) highly virulent procaryotes can acquire iron by siderophore-independent processes and may utilize hemin as the _sglg source of iron (101); (iv) procaryote facultative intracellular parasites can penetrate nonprofessional phagocytes (14,54,80,132); and (v) ability to absorb exogenous hemin or the dye Congo red serves as a virulence factor (77,125). Results of these and other findings, all of which later proved applicable to other pathogenic species, have now established the yersiniae system as the most important basic experi— mental model for establishing mechanisms of virulence in procaryote facultative intracellular parasites. An advantage of the yersiniae model is the availability of several conditionally virulent strains of 11:. pestis which permit the safe handling of organisms in the laboratory. The individual virulence determinants of 1. m have been discussed in detail by Burrows (35) and Brubaker (19). Following a brief summary of the effects of iron privation upon the expression of virulence by procaryotes is an updated review of the virulence determinates of yersiniae. 5 TABLE 2 . Plasmids in yersiniae (26,58,83). Plasmid (megadaltons ) Phenotype Species 6 Pesticinogeny (Pst+) l. Estis Fibrinolysin/Coagulase Pesticin immunity 45 Low-calcium response (Lcr+) 1. Estis yersiniae outer-membrane _Y_. pseudotuberculosis peptides (Yops) V and W virulence antigens l. enterocolitica 65 Unknown 1. Estis ? Murine toxin IRON PRIVATION HOST DEFENSE Once a potentially invasive bacterium has bypassed dermal barriers, it becomes vulnerable to destruction by a series of specific and nonspecific host extracellular antibacterial systems including complement (15,128), B-lysin (55), and lysozyme (116). Iron privation in mammalian extracellular fluids also serves as a major defense mechanism against bacterial growth (134,136,137). Although serum and tissue contain iron in excess of the amount required for bacterial growth, iron is associated with strong humoral iron-binding proteins and is not readily available for use by invading microbes (137). The 3+ in the plasma, present as a already low concentration of Fe transferrin-Chelate, is even further reduced by intracellular deposi- tion during the febrile response. Bacteria within the bloodstream thereby encounter an essentially iron-free environment (except for that present in hemoglobin of erythrocytes whidh is generally unavailable to procaryotes) (136,137). To initiate growth in the body of an animal, any potential pathogen must first be able to obtain sufficient iron for its metabolism. Iron is probably essential for the growth of all cells, especially the ones we are likely to encounter as potential pathogens. Iron is required for a variety of enzymatic reactions in which the cation serves as a cofactor per se or as part of a prosthetic group, primarily heme. Sublethal degrees of iron-privation cause 7 bacteriostasis by inhibiting these reactions, whereas, severe degrees of iron-privation are lethal due to lesions at the level of protein synthesis. Studies in g. ggli_(127,138) indicate that iron is necessary as a cofactor in the enzymatic synthesis of 2-methy1thio groups needed for the function of tRNA molecules which have uracil as their first codon letter. Due to conservation of the tRNA structure utilized in translation, it is possible that all organisms will require at least trace levels of iron for growth. IRON ACQUISITION Two major mechanisms used by bacteria to acquire iron are (1) Chelation with transport by siderophores and (2) uptake mediated by specialized surface structures. Siderophores are soluble, low molecu- lar weight phenolate or hydroxamate microbial iron carriers (89,97) that are probably necessary for growth in many natural environments. Once widely assumed to be involved in the expression of virulence, results now show that mutational loss of siderophore production by salmonellae (8) and Pseudgmonas (1) did not significantly reduce the ‘virulence of those organisms. Results obtained with yersiniae showed that a highly pathogenic microbe could obtain iron via a siderophore- independent high—affinity cell-bound system (101). Subsequent work by others indicated that neisseriae (2,118), listeriae (50), and legionellae (112) possessed simfilar'siderophore-independent iron-uptake medhanisms. Whether the previously mentioned facultative intracellular parasites utilize intracellular storage reserves such as ferritin is unknown. Yersiniae are capable of utilizing hemin, a constituent of many host ligands, as a sole source of iron (101). 8 BACTERIOCIN AND SIDEROPHORE RECEPTORS Bacteriocins of Escherichia coli, termed colicins, are separated into groups A and B, depending upon the nature of their corresponding outer—membrane absorption site (52,53). Receptors of group B colicins are typically involved in iron metabolism. For example, mutational loss of the structure required for uptake of the endogenous phenolate siderOphore enterochelin (ER) or that for exogenous hydroxymate siderophores (_f__h;i_A_) promotes resistance to colicins B or D and colicin M, respectively (52). Pesticin (6) is an analogous bacteriocin produced by wild type _11. Estis (Pst+) known to inhibit certain strains of E. go_l_i_, X. pseudotuberculosis, X. enterocolitica, and Pst- isolates of Ir< . Estis retaining the pigmentation phenotype. The nature of the pesticin receptor is not yet known. However, three findings outlined below suggest that this structure, like those for group B colicins, may be required for assimilation of iron. First, lethality of pesticin was 3+ or hemin, suggesting that its receptor is prevented by exogenous Fe repressed by the cation (26). Second, although resistance to pesticin (Pstr) was not shared by gap, £13315, or _c_i_§ mutants of E_.. _co_li d, to_n_l_3_ isolates of this organism were also tolerant to the bacteriocin (58). Finally, mutation to Pstr in Pst‘ isolates of X- Estis which lack immunity to the bacteriocin (28) resulted in avirulence due to concomitant conversion to Pgm- (18). PURINE BIOSYN'THISIS The ability to synthesize purines gg ggyg_(Pur+) is an estab- lished virulence determinant of all pathogens. The first report that mutation to purine auxotrophy resulted in avirulence was shown for Salmonella typhgsa (3). Similar reports followed for Klebsiella pneumoniae (63), Yersinia pestis (31), Pseudomonas pseudomallei (90), and Bacillis anthracis (76). It is believed that free purines are unavailable within the mammalian host and that a metabolic block in purine biosynthesis results in complete avirulence (3). Brubaker later showed that if this block prevents the synthesis of inosine monophOSphate (IMP), virulence is only slightly reduced; if the blockage occurs subsequent to the synthesis of IMP, virulence is lost completely (17). FRACTION I The envelope of Yersinia pestis contains a protein-carbdhydrate complex (Fraction I) (Fra+) that is produced mainly at 37Gb, confers antiphagocytic properties, and activates complement (59,81). Fra— cells were unable to synthesize the fraction I capsular antigen (36, 104), whereas FI:t cells (34) produced fraction I antigen but were unable to incorporate the antigen into the extracellular matrix. 10 The mutation to Fra" may occur at high frequency (104), as shown in continuous culture. Fra- cells remained fully virulent for mice 3 when injected intraperitoneally (19), but resulted in a 10 to 106-fold increase in ID for guinea pigs injected similarly (33) . However, 50 retention of virulence was shown for Fra- cells via intradermal injec- tion in guinea pigs. The importance of fraction I antigen, determined by immunoelectrophoresis is in doubt with regard to pathogenicity for humans (139). ICRPIAMD Virulence of all yersiniae is dependent upon the presence of a 72-kilobase low—calcium response (Lcr+) plasmid (7,25). At 37°C in Ca2+ deficient medium, this plasmid promotes restriction of growth with concomitant production of virulence functions including V antigen and a specific set of yersiniae outer—membrane peptides termed Y0ps (11 ,12, 120). In Yersinia Estis both restriction and synthesis of V antigen are potentiated in this environment by elevated m2+ and prevented by Ca2+ or exogenous nucleoside triphosphates (12,29,145). It is well known that Ca2+ is required for many functions of eukaryotic organisms . However , vegetative growth of prokaryotes rarely requires the presence of Ca2+ (21,103,144). This general assumption caused a considerable delay in research progress and in the clinical isolation of virulent yersiniae. Many types of bacteriological media were tried, yet rapid loss of virulence occurred when cultures were cultivated at 37°C (62). 11 Studies by Kupfenberg and Higudhi (84) and then Higuchi gtual. (69) showed that virulent cells of X, pestis required at least 2+ 2.5 mM Ca in order to grow at 37°C; the cation was not needed for multiplication at room temperature. These workers acknowledged that 2+ or Zn2+ also could promote similar concentrations of either Sr division; only Ca2+ exists at a sufficient concentration within the host to permit cell divisions ig_vivo. Later work by these groups showed that Ca2+ dependency was maximal in the presence of intracel- lular levels (mammalian) of 1492+ (20-40 m), and Sargalla (124) first showed that the magnesiumedependent requirement for Ca2+ was lost when cells became Lcr_. In 1960, Burrows and Bacon (40) showed the existence of V antigen in virulent isolates of Yersinia pseudotuberculosis. Carter gt 2;, (45) made the same observation for Yersinia enterocolitica, a recent addition to the genus Yersinia. The reproduceability of growth restriction requires adaptation of 2 the bacterial cells (eight to 10 doublings) to the Ca +-deficient medium and simulation of an intracellular environment with regards to M92+ concentrations (20mM). Shifting incubation temperature from 26°C to 37°C results in a sequence of events known as the low-calcium response (65). Postshift cells within the population were shown to accumulate sufficient DNA to account for one doubling in cell mass in addition to rounds of replication already in progress at the time of the shift (91). The first event observed during restriction was the shutoff of stable RNA synthesis, but not necessarily messenger RNA synthesis, whiCh was followed by a decrease in adenylate energy charge and then by cessation of DNA and protein synthesis (48,144). This 12 ordered metabolic stepdown was shown to be independent of the regula- tory nucleotides guanosine tetraphosphate or guanosine pentaphosphate and probably represents a primary block in stable RNA synthesis (48). The Lcr plasmid also encodes at least 11 outer-membrane proteins (Yops) (see Table 3) whidh are expressed maximally during growth at 37°C in the absence of Ca2+. Originally seen only in X} pseudotuberculosis and X, enterocolitica (120,121), there was evidence using convalescent sera Obtained from recovered plague patients that X} p§§§i§_also synthesized Y0ps. By transferring the Lcr plasmid of X. pgsgg to X. geudotuberculosis, it was shown that xg‘p§§5i§_also had the capability to express Y0ps as major constituents of the outer-membrane (140). Yop expression is stable in the enteropathogenic species but not by wild type 19 p§§§i§_whiCh possesses a unique 10 kb Pst plasmid associated with pesticinogency (114). Pulse-Chase labeling experiments show that Y0ps are equally expressed by both Lcr+, Pst" and Lcr+, Pst+ cells of X} E§§Ei§j however, they are rapidly degraded by a protease encoded on the pesticin plasmid of the latter strain (115). These findings suggest that a product of the Pst plasmid of X. pgs_1_:_i_s_ is required for post-translational regulation of outer-membrane peptides mediated by a second unrelated Lcr plasmid. The role of prs in promotion of virulence is currently contro- versial. Subsequent work in this area will prove to be important in the elucidation of additional virulence medhanisms of yersiniae. 13 TABLE 3. Accepted nomenclature and approximate molecular weights of Yops expressed by the yersiniae.a Accepted Alternate Approximate nomenclature nomenclatureC mol. weightd YopA Yopl 150-200 kdal YopB Yop2 44 kdal YOpC Yop3 40 kdal YopD Y0p4 34 kdal YopE Yop5 26 kdal YopF NOne 76 kdal YopG NOne 61.5 kdal YopH ane 49.2 kdal YopI ane 46.3 kdal YopJ NOne 31 kdal YopK ane 21 kdal All Yops reported to be produced by the yersiniae. Net all Yops are produced by all species or strains. meenclature is taken from Straley and Bowmer (119). Taken from Bolin et al. (11,12). As estimated by SDS—polyacrylamdde gel electrophoresis. 14 PESTICIN PLASMID Cells of wild type Yersinia pestis encode a bacteriocin (6) termed pesticin and the invasive enzymes coagulase and fibrinolysin on a common 10-kilobase plasmid (Pst+) (5,30,57). Loss of the pesticin plasmid results in avirulence by intraperitoneal or subcutaneous routes of injection, but not by the intravenous route (26,126). Full restor- ation of virulence is adhieved by simultaneously injecting the experi— mental animal with sufficient iron to saturate serum.transferrin levels (17). Pesticin, a 65,000 dalton protein, exhibits an N-acetylglucosaminidase activity (56) whidh converts sensitive cells to osmotically stable spheroplasts (67). Sensitivity to pesticin is apparently limited to the genus Yersinia and certain strains of ESCherichia coli. Serotype I strains of X} pseudotuberculosis (40), some serotype 0:8, 0:21, and 0:4,32 isolates of X, enterocolotica (58, 102), the universal colicin indicator strain E, 921; (58,61), and non-pesticinogenic mutants of X} p§§t1§_whiCh retain the Pgmf phenotype are all sensitive to pesticin. Wild type cells of X, pestis, that produce pesticin, putatively produce an immunity protein whidh renders them insensitive to the effect of pesticin; it has not yet been identified. Retention of the pesticin plasmid is assured within the p0pulation as the loss of pesticinogeny would result in concomitant loss of pesticin immunity with subsequent cell destruction. Thus, 15 expression of this virulence factor by all members of the bacterial population is assured. Pesticin activity is maximal at pH 4.7 and is 20 times more active at 37°C than at 36°C (56). Brubaker and Surgalla noted that the effect of pesticin can be inhibited by hemin and Fe3+. This effect can be reversed by Ca2+, Sr2+, or chelating agents (such as ethylene-diamine 2+ which reduces the available iron concen- tetraacetate) in excess Ca tration within the environment. Selection of 50 pesticin resistant mutants (Pstr) from Pgmf, Pst- strains of X, pestis resulted in concomitant conversion to Pgm-. Likewise, 50 nonpigmented mutants selected from the same strain were all found to be pesticin resistant. USing this relationship, the 5 mutations per mutation rate from Pgm+ to Pgm- was determined to be 10- bacterium per generation (18). The selection of Pstr mutants of enteropathogenic yersiniae also resulted in isolates of reduced virulence via peripheral routes of injection (132); similar to Pgmf mutants, the decreased retention of bacteria within host internal organs allowed recovery by the experimental animal unless sufficient iron was administered intraperitoneally to saturate humoral iron binding proteins (78,132). A common mechanism of virulence was postulated for yersiniae expressing a putative pesticin receptor. In addition to its role in the invasiveness of X, pestis, the plasmin activator encoded on the pesticin plasmid is responsible for the post—translational regulation of outeramembrane Yops but not soluble peptides mediated by a second unrelated Lcr plasmid (114,115). 5 After pulsing with 3 Semethionine, peptides with molecular weights corresponding to all known Yops are synthesized during the low calcium 16 response by both Lcr+ , Pst+ and Lcr+, Pst- cells of 1. Eggs, yet radioactivity in Yops of wild type was rapidly chased into lower molecular weight degradation products (115). A major area of study currently relates to the four soluble peptides that are expressed in a stable manner during the low calcium response. PIQIENTATION Pigmentation in _Y_. pestis was first described in 1956 by Jackson and Burrows (77) as the ability to absorb exogenous hemin from a defined solid medium and thus grow in the form of colored or pigmented colonies (Pgm+). Mutational loss of this function resulted in growth as white colonies with a marked reduction of virulence in mice by intraperitoneal (78) or subcutaneous (132) routes of injection. Surgalla and Beasley (125) later developed a common laboratory medium that utilized Congo red to differentiate Pgm+ from avirulent Pgm— cells. Modifications of this latter medium were later shown to be useful in identification of virulent isolates of Escherichia coli, y_i_b;_i_g and Shigella species and Neisseria meningitidis (9,47,93,100). Jackson and Burrows (78) showed that avirulent Pgm- cells can be restored to full virulence in mice upon injection of sufficient iron to saturate serum transferrin, thus providing an excess of iron in the hosts' plasma. This evidence implied that Pgm+ cells possess the ability to acquire iron i_n_ y__i__v_o_ from a source not available to Png mutants. later studies by Jackson and Morris (79) stated that neither Pgm+ nor Pgm’ cells were capable of growing in mouse serum unless an 17 excess of iron was added. In addition, supplementation with hemin from lysed erythrocytes failed to stimulate growth in mouse serum. Whether the added iron has an effect on the component of the serum necessary to kill the cells, or the Pgm+ cells are capable of using a source of iron igLyiyg_that is unavailable to Pgmr mutants is unknown at this time. The frequency for spontaneous mutation from Pgm+ to Pgm. is reported to be 10‘5 mutations per generation per bacterium (18), a very high rate for a reaction not shown to be associated with an extra- chromosomal element (57,130). So far the reverse mutation of Png to Pgm+ has not been detected (37). It has been suggested that mutational loss of pigmentation is the result of a chromosomal deletion; the number of gene(s) involved has not been determined. Perry and Brubaker (101) showed that both Pgm+ and Png cells of _l_(_. Estis were capable of identical growth rates in a defined medium rendered highly iron-deficient (10,135) , and that hemin could serve as the £915 source of iron. Additionally, it was shown that yersiniae can acquire iron by an inducible, high-affinity, cell-bound, siderOphore- independent process (101) . These observations were rapidly extended to include many highly virulent facultative intracellular pathogens including the neisseriae (2,118), listeriae (50), and legionellae (112). Mutational loss of siderOphore production by salmonellae (8) and Pseudomonas (1) did not significantly influence the expression of virulence by these organisms but may be essential for survival in most natural environments. It is now generally believed that facultative intracellular parasites acquire iron via unique siderophore—independent processes . 18 Enteropathogenic yersiniae are found to be Pgm' when grown under conditions used to define pigmentation in X- pestis (125). Sensitivity to pesticin, a phenotype associated with Pgm+, Pst' cells of 1. pe_s_tg_s_, is found in serotype I strains of X. pseudotuberculosis and certain serotypes of X. enterocolitica as mentioned previously. Combined with the emergence of one Pgm+, Pst- strain of l. pestis that is insensitive to pesticin (19), this information suggests that the absorption site for pesticin is not necessarily the same as for hemin. 19 TABLE 4. Virulence factors of Yersinia pestis and their effect on the LD50 for mice via various routes of infection.a Phenotype of X. Estisb Route of infectionC Pur Fra Pgm Pst Lcr s.c. i.p. (i.p. + Fe) i.v. + <101 <101 <101 <101 - 108 1o8 NRd 108 + NR <10l NR NR + >107 >107 <101 1.5 x 101 + >107 105 <101 7.1 x 101 + + + + - >10.7 >107 >107 >107 a. Data taken from Brubaker (19), Brubaker, Beesley, and Surgalla (26) and Une and Brubaker (132). b. Abbreviations used are as follows: Pur, purine biosynthesis Fra, expression of Fraction 1 capsular material Pgm, the ability to absorb exogenous hemin Pst, carriage of the pesticin plasmid Lcr, carriage of the low calcium response plasmid c. Abbreviations used are as follows: s.c., subcutaneous injection i.p., intraperitoneal injection i.v. , intravenous injection d. NR stands for not reported in the literature. 20 CHAPTERI (ARTICLE) Resistance to pesticin, storage of iron, and invasion of HeLa cells by yersiniae bY Daniel J. Sikkema and Robert R. Brubaker Published in Infection and Immunity Volume 55, pages 572-578 INFECTION AND IMMUNITY. Mar. 1987. p. 572-578 0019-9567/87/030572-07502.00/0 Copyright (0 1987. American Society for Microbiology 21 Vol. 55. No. 3 Resistance to Pesticin, Storage of Iron, and Invasion of HeLa Cells by Yersiniael DANIEL J. SIKKEMA AND ROBERT R. BRUBAKER‘ Department of Microbiology and Public Health. Michigan State University. East Lansing. Michigan 48824-110! Received 21 August l986/Accepted 17 November 1986 The independent abilities of Yersinia pestis to absorb exogenous pigments including hemin and Congo red (Pgm*) and to produce the bacteriocin pesticin with genetically linked invasive enzymes (PstI) are established virulence factors of the species. Pst" Pgm’ strains of Y. pestis are sensitive to pesticin (Pst'). and mutation of these isolates to pesticin resistance (Pat) is known to result in concomitant conversion to Pgm‘. Wild-type cells of Yersinia pseudotuberculosis and Yersinia enterocolitica are Pgm" but may be Pst'; mutation of the latter to Pst' also results in avirulence. In this study, typical Pgm" mutants of Y. pestis exhibited a dramatic nutritional requirement at 37°C but not 26°C for iron which could be fulfilled by either Fe” or hemin. Iron privation of Pgm" yersiniae resulted in formation of osmotically stable spheroplasts similar to those previously observed after exposure of Pst“ bacteria to pesticin. At 37°C, 1’ng organisms rapidly overgrew initially predominant Pgm“ populations in iron-deficient medium. However, Pgm" isolates could undergo a second mutation that permitted successful competition with Pgm” cells in this environment. The mutation to Pst' in Y. pseudotuberculosis and Y. enterocolitica did not promote a similar requirement for iron but rather prevented these organisms from penetrating HeLa cells. The ability to invade these nonprofessional phagocytes was not shared by Pgm+ or Pgm' cells of Y. pestis. Wild-type cells of Yersinia pestis. the causative agent of bubonic plague. absorbed exogenous hemin (26) or the dye Congo red (43) during growth at 26°C on solid media thereby forming intensely colored or pigmented colonies (Pgm’). Mutational loss of this evident chromosomal function (17. 42) resulted in comparable growth as white colonies (26. 43). disappearance of a major outer membrane peptide (42), and marked reduction of virulence in mice by intraperitoneal (27) or subcutaneous (47) routes of injection. Typical cells of closely related Yersinia pseudotuberculosis and Yersinia enterocolitica were Pgm" as judged by their growth as white colonies under conditions used to define pigmentation in Y. pestis (43; R. R. Brubaker. unpublished observations). Sub- sequent work showed that other pathogenic species includ- ing neisseriae. shigellae. Vibrio, and certain isolates of Escherichia coli were Pgm+ as judged by their ability to absorb Congo red in environments distinct from that estab lished as optimal for Y. pestis (5. 13. 33. 35): this determinant was correlated with the ability of shigellae to penetrate nonprofessional phagocytes (13). Ability to produce the bacteriocin pesticin (2. 23. 24) and genetically linked (11. 31) coagulase and fibrinolysin (1) serves as an additional virulence factor (9) of Y. pestis (Pst’). Loss of these functions. which are mediated by an approximate 6 megadalton plasmid (3. 17. 42). also promoted avirulence in mice by intraperitoneal and subcutaneous routes of injection (9). Y. pseudotuberculosis and Y. entero- colitica cells are Pst‘. but isolates of certain serotypes of these species (12. 36) and a few strains of E. coli (10. 18. 24) were also sensitive to the antibacterial action of pesticin (Pst’). This lethal efl'ect was catalyzed by the N-acetylglu- cosaminidase activity of the bacteriocin (16) which promoted conversion of Pst‘ bacteria to osmotically stable sphero- plasts (20). Expression of Pst‘ in Y. pestis required loss of ' Corresponding author. 1’ Article no. 12071 from the Michigan Agricultural Experiment Station. 572 immunity to the bacteriocin by mutation to Pst' (21) with retention of the 1’ng determinant (8). Subsequent mutation of these Pst' Pgm’ organisms to pesticin resistance (Pst') resulted in selection for Pgm“ isolates thereby permitting determination of a mutation rate from Pgm' to Pgm" of 10‘ (8). Analogous Pst‘ mutants of Y. pseudotuberculosis and Y. enterocolitica were subsequently isolated and. like Pgm“ and Pst‘ cells of Y. pestis. were of reduced virulence in mice via intraperitoneal and subcutaneous routes of injection (47). Pgm’ and Pst’ yersiniae were maintained in mouse liver. spleen. and lung. whereas Pgm' and Pst" mutants were rapidly cleared from these organs. suggesting loss of a common mechanism of virulence (47). The Pgm+ phenotype was assumed from the outset (7) to favor accumulation of iron in vivo (26. 27). However. attempts to demonstrate a selective advantage of Pgm’ organisms in iron-deficient environments including serum (28) or 8-hydroxyquinoline- extracted medium (37) were not successful. Cells of Y. pseudotuberculosis, Y. enterocolitica. and both Pgm’ and Pgm‘ Y. pestis utilized hemin as a sole source of iron and also accumulated Fe” by an inducible. siderophore- independent. cell-bound. high-affinity transport system (37). This finding suggested that all these organisms acquire iron by the same mechanism and that the Pgm’ determinant merely serves as a mechanism to store the cation. The purpose of this report is to provide evidence indicating that avirulence associated with mutation to Pgm“ in Y. pestis is indeed correlated with a significant lesion in iron accumula- tion. in contrast. avirulence caused by the analogous muta- tion to Pst‘ in Y. pseudotuberculosis and Y. enterocolitica reflected loss of ability to invade nonprofessional phago« cytes. MATERIALS AND METHODS Bacteria. All yersiniae used in this study were avirulent owing to loss of the ~45-megadalton plasmid that mediates the low-calcium response (3. 17). Unless stated otherwise. VOL. 55. 1987 22 PESTICIN RESISTANCE IN YERSINIAE 573 TABLE 1. Properties of yersinia strains . . . . . Plasmids Biotype’ Species Strain “‘23:: 30:21:23 $22323“ tmega- or sero- Source or origin daltons) type Y. pestis Kuma + + l 6. 65 A Manchuria KIM + + l 6. 65 M Iran G32 + - S 65 0 United States TS + + I 6. 65 0 Java M23 + + l 6. 65 O Nonencapsulated mutant of MP6 A12 + — S 65 O From strain A1122 (United States) Salazar + + l 6. 65 0 Bolivia Yokohama + + l 6. 65 A Japan Kimberley + + l 6. 65 0 Africa Dodson + - S 65 0 Unknown K10 + + l 6. 65 M Iran MP6 + + I 6. 65 0 Unknown Y. pseudotuberculosis PBI - - 5 None 1 England 1 - - S None 1 Unknown M067 - — 8 None I United States Hale -— - 8 None l England Galligue - - 8 None I England Parkin — - 8 None 1 England Y. enterocolitica Wa - - S None 0:8 United States P76 - - S None 0:8 United States E701 - - S None 0:9 Canada E736 - - S None 0:21 Canada “ S. Sensitive; I. immune. ° Biotypes are antique (A). mediaevalis (M). and orientolis (O). organisms used in experiments were Y. pestis Kuma (37). Y. pseudotuberculosis P81 (12). and Y. enterocolitica WA (36). Salient properties of these strains and other yersiniae are shown in Table 1. Pgm' mutants of Y. pestis were obtained on Congo red agar (43) incubated at 26°C. and Pst' mutants of Y. pseudotuberculosis and Y. enterocolitica were isolated on solid medium containing homogeneous pesticin (23) by the method previously described for determining the muta- tion rate of Y. pestis to Pgm" (8). Streptomycin-resistant mutants were selected at 26°C for the ability to form colonies on tryptose-blood agar base (Difco Laboratories. Detroit. Mich.) containing the antibiotic (100 U/ml). Cultivation. Slopes of tryptose-blood agar base were incu- bated for 1 day (Y. pseudotuberculosis or Y. enterocolitica) or 2 days (Y. pestis) at 26°C after inoculation from stock cultures of buffered glycerol maintained at -20°C as previ- ously described (1). Resulting organisms were suspended and appropriately diluted in 0.033 M potassium phosphate buffer (pi-i 7.0) (phosphate buffer) for use as inocula of chemically defined medium (S3).°yielding an initial optical density at 620 nm of 0.1 (corresponding to about 10‘| viable bacteria per ml). A single transfer was made in this medium containing 50 aM Fe” or 20 to 50 aM hemin in Erlenmeyer flasks (10% [vol/vol]) aerated at 200 rpm on a model 076 gyratory water bath shaker (New Brunswick Scientific Co.. lnc.. Edison. NJ.). Organisms in late logarithmic growth were harvested by centrifugation (17.000 x g for 10 min at 4°C). washed once in phosphate buffer. and then suspended at an optical density of 0.1 in control medium containing iron as described above or in iron-deficient medium. These cultures were aerated as already defined and. in some experiments. were used to inoculate subsequent transfers when the optical density had increased to 1.0. Iron-deficient medium was prepared by extraction with 8-hydroxyquin- oline (50) and contained <0.3 aM iron as determined by flame absorption spectroscopy. Host-cell invasion. HeLa cells were prepared in suspension and used in experiments with yersiniae essentially as previ- ously described (15). The cells were routinely cultivated at 37°C under 5% CO; in Eagle minimal essential medium (Difco) with 5% fetal calf serum containing streptomycin (75 jig/ml) and potassium penicillin G (100 U/ml). After growth to a density of 5 x 10’ cells per ml. antibiotics were removed by three washes in minimal essential medium with fetal calf serum. and 2.0 ml of the final suspension was placed in roller tubes. The latter then received 10" yersiniae washed once in phosphate buffer in 0.1 to 0.2 ml of the same butfer and were placed in a model TC~5 roller apparatus (New Brunswick) at 37°C. After infection for 2 h. extracellular organisms were killed by further incubation in the presence of kanamycin ( 75 pig/ml) for 4 h. Intracellular bacteria were then released by disruption of the HeLa cells with a Dounce homogenizer and determined by plating appropriate dilutions on tryptose- blood agar base. Results of preliminary studies showed that significant uptake. when it occurred. commenced after 1 h and ceased after 2 h of incubation (15). RESULTS The degree of adaptation to the culture medium and the extent of prior exposure to iron markedly influenced bacte- rial growth during subsequent iron privation. To control these variables. the organisms were cultivated for a single transfer with Fe“° or hemin before washing and dilution for use as inocula. Pgm’ cells of Y. pestis were capable of evident unlimited growth thereafter at 37°C in iron-deficient 574 SIKKEMA AND BRUBAKER 10.0 1.0 i >- : t 4 w 1 z tu 0.1 ‘3 fifi‘ - f - a 4 e J F o < 9100£ 9 ' “3:1 / E 3 f . s O i if i f 1% 1 1‘0L is is l at 'i . ': it 1: 1 t it 37‘ 1: : 1 2:1 it! I: ' 3 I‘i i ‘l 0'1: ii! - LIE! A! . 1 0 18 360 18 360 18 360 18 36 HOURS F IG. 1. Optical density of Pgm’ cells of Y. pestis Kuma during the first (A). second (B). third (C). and fourth (D) transfer and of Pgm‘ mutants during the first (E). second (F). third (0). and fourth (H) transfer at 37°C in iron—deficient medium (0) or medium supplemented with 50 pM FeCl, (0). The organisms were initially grown with 50 nM Fe". washed. and inoculated into the first transfers; subcultures were inoculated when the Optical density of the parent culture was 1.0. Pgm‘ mutants failed to grow after the first transfer. medium when prepared by this process after an initial transfer with Fe3+ (Fig. 1) or hemin (Fig. 2). Pgm’ organisms also exhibited comparable generation times at 26°C in iron- deficient medium or in media supplemented with Fe” or hemin. although the latter promoted marked autoagglutina- tion (data not shown). In contrast. Pgm“ mutants exhibited only limited multiplication at 37°C upon transfer to iron- 5 O "" in ""713? .. - .1 0'1 4 1.. - O 1... ITEV‘I“ ' I'll. OPTICAL DENSITY O A A ALAA 1 1 4 1 ‘1 1 4 w 1 4 1.0g ‘3 i 1 ;, 2 . i l 1 wave.» f r l 0.1"” - as’ . - 0 18 380 18 380 18 330 18 36 HOURS FIG. 2. Same as Fig. 1 except that 50 uM hemin was used in place of Fe”. 23 th-‘ecr. IMMUN. 'TA a ’. B a >100.— if. 1 l- e’l» e. g L v' 1; , i s l H ' ] 10r a W 3 15°: . ° 3 E ; °,° t paw 00:01- O . ‘r.. i Q .7 J 0.1 O 25 500 25 50 HOURS FIG. 3. Optical density at 26°C (0) or 37°C (C) of Pgm" mutants of Y. pestis KIM in iron-deficient medium after a previous single transfer in the same medium at 26°C (A) or 37°C (Bl. deficient medium after saturation of storage reservoirs with Fe” (Fig. 1) or hemin (Fig. 2). As shown below. growth of Pgm' organisms at 26°C in iron-deficient medium was equiv- alent to that observed at 37°C with added iron. Pgm" mutants were cultivated for one passage with Fe" at 37°C and then subcultured once in iron-deficient medium at either 26 or 37°C. Regardless of the prior temperature of cultivation. a further transfer in iron-deficient medium at 26°C resulted in a lag followed by full-scale growm (Fig. 3). In contrast. similar incubation at 37°C resulted in restriction of growth. although significant residual multiplication oc- curred in the culture inoculated with organisms previously grown at 26°C. Accordingly. Pgm‘ mutants specifically exhibit an evident lesion in accumulation of iron at 37°C but not at 26°C. Although the nature of this temperature dependent deficiency is unknown. it provides Pgm’ bacteria in iron-deficient medium with a marked selective advantage. For example. an inoculum consisting of an artificial mixture of 99.95% Pgm” and 0.05% Pgm’ organisms underwent a rapid and dramatic shift to an essentially pure Pgm’ popu- lation during a single transfer in this environment (Fig. 4). These and related studies showed that the viability of Pgm‘ mutants at 37°C often decreased more rapidly in iron- deficient medium than did the optical density. This discrep« ancy was resolved upon observation of iron-deprived Pgm ' 10F f r 2 _l 9* /4. E y a, 87- 1 _I d 7 . o L“, 6—0 e l m S 5» « > 1 (9 4 3 4 a». - . A 0 20 4O 60 HOURS FIG. 4. Total cells (0) and Pgm’ cells (C) of an artificial mixture initially composed of 99.95% streptomycin-sensitive Pgm’ mutants and 0.05% streptomycin-resistant Pgm’ organisms determined on tryptose-blood agar base and streptomycin agar. respectively. VOL. 55. 1987 g r 1'. \ .e ’ 0L) '1‘ (”I i .’ ’>. '7 FIG. 5. Pgm’ cells of Y. pestis Kuma exhibiting normal mor- photogy (A) and osmotically stable spheroplasts of Pgm‘ mutants (B) after a single transfer in iron-deficient medium. and Pgm‘ organisms by light microscopy. The former ex- hibited normal morphology. whereas the latter underwent conversion to nonviable osmotically stable spheroplasts that nevertheless contributed to optical density (Fig. 5). Other strains of Y. pestis were examined to determine whether the Pgm‘-specific lesion in iron metabolism was a general property of the species. Of the 11 additional isolates tested. 8 resembled strain Kuma in that only Pgm' organ- isms were capable of growth at 37°C without added iron (Table 2). Pgm“ mutants of the remaining three strains grew as well in this environment as did their Pgm“ parents. This finding suggested that Pgm“ bacteria typically exhibit the temperatureodependent defect since the exceptions pos- TABLE 2. Doubling times and maximum Optical density of I’grn° and Pgm’ strains of Y. pestis at 37°C in chemically defined iron- deficient medium" . Maximum Strain Pgm WNW ' time (min) density Kurna + 110 8.160 - NG‘ 0.125 KIM + 110 6.120 - NG 0.122 032 + 150 7.960 - NG 0.137 TS + 240 5.950 - NO 0.129 M23 + 170 6.340 - 180 6.200 A12 + 220 7.000 - NO 0.119 Salazar + 170 7.740 - NO 0.141 Yokohama + 220 6.300 - NO 0.126 Kimberley + 240 4.300 - 240 5.550 Dodson + 210 8.070 - 230 7.080 K10 + 150 6.800 - NO 0.095 MP6 + 270 2.570 - NO 0.084 ‘ Organisms were inoculated at an optical density of 0.1 after a single transfer in iron-deficient medium. ° NC. No growth (doubling time >24 h). 24 PESTICIN RESISTANCE IN YERSINIAE 575 sessed various atypical characteristics and were of uncertain background (Table 1). However. the existence of these exceptions demonstrated that the ability to express the pigmentation reaction was not necessarily essential for growth at 37°C in iron-deficient medium. To further illustrate this point. Pgm‘ cells of strain Kuma were cultivated at 37°C for 4 days without added iron at which time a subpopulation of otherwise typical Pgm“ organisms emerged that were capable of evident indefinite growth in this environment. These observations indicate that a suppressor mutation can promote multiplication of Pgm" cells at 37°C in iron-deficient medium. Thus. acquisition of this ability does not reflect true reversion to Pgm’. Since Pgm‘ mutants are analogous to Pst' mutants of Y. pseudotuberculosis and Y. enterocolitica. the latter were compared for ability to grow at 37°C in iron-deficient me- dium. Control Pgm' cells of Y. pestis typically ceased division in this environment (Fig. 6A). whereas results obtained with Y. pseudotuberculosis were equivocal in that the Pst' mutant grew more slowly than did the Pst" parent but eventually achieved the same maximum optical density (Fig. 68). No difi'erence was detected between Pst’ and Pst' cells of Y. enterocolitica (Fig. 6C). These findings indicated that mutation to Pst' in Y. enterocolitica and probably Y. pseudotuberculosis did not result in significant loss of ability to accumulate iron as occurred with Pgm‘ mutants of Y. pestis. Accordingly. a search was initiated for other putative virulence functions present in Pst‘ but not Pst' mutants of Y. pseudotuberculosis and Y. enterocolitica. Attempts to dem- onstrate difi'erences in adherence or ability to damage host cells were not successful. but the capability to penetrate nonprofessional phagocytes was significantly reduced in Pst’ mutants. This relationship is shown in Table 3 which shows that mutation of Y. pseudotuberculosis and Y. enterocolitica to Pst‘ resulted in significant decreases in the ability to invade HeLa cells. Neither Pgm’ nor Pgm‘ isolates of Y. pestis were capable of promoting detectable invasion of HeLa cells. DISCUSSION It is known that macrophages serve as target host cells for Y. pestis. Y. pseudotuberculosis. and Y. enterocolitica (29. 44. 46). This relationship is emphasized by the observation that organisms of all three species are of highest virulence FIG. 6. Growth of Pgm' (O) and Pgm’ (0) cells of Y. pestis Kuma (A). Pst’ (O) and Pst' (0) cells of Y. pseudotuberculosis [’81 (B). and Pst’ (O) and Pst' (0) cells of Y. enterocolitica WA (C) during a second transfer in iron-deficient medium. 576 SIKKEMA AND BRUBAKER via intravenous injection (9. 47). in which immediate inter- action occurs with fixed macrophages lining the blood ves- sels of the liver and spleen. Further evidence underscoring the importance to yersiniae of obtaining access to macro- phages was the observation that virulence of Pgm' and Pst‘ mutants of Y. pestis (9. 47) and Pst' mutants of Y. pseudotuberculosis and Y. enterocolitica (47) could be phe- notypicaliy restored via intravenous injection. This finding suggests that these mutants are rapidly eliminated from the host by nonspecific mechanisms of defense after injection by peripheral routes but are able to cause acute disease after administration by a route that permits immediate uptake by macrOphages (47). An important clue to the nature of these nonspecific mechanisms of host defense was the discovery that virulence of Pgm". Pst', and Pstr mutants could be fully restored via peripheral routes of infection by saturation of serum transferrin by injected iron (9. 27; unpublished obser- TABLE 3. infectivity for HeLa cells of Pgm’ and Pgm‘ isolates of Y. pestis. Pst' and Pst’ isolates of Y. pseudotuberculosis. and Pst‘ and Pst' isolates of Y. enterocolitica Species Strain Phenotype Infectivity‘I Y. pestis Kuma Pgm’ <1 Pgm’ <1 KIM Pgm’ <1 Pgm‘ <1 G32 Pgm’ <1 Pgm ' <1 TS Pgm’ <1 Pgm" <1 M23 Pgm° <1 Pgm ' <1 A12 Pgm ’ <1 Pgm' <1 Salazar Pgm ° <1 Pgm‘ <1 Yokohama Pgm’ <1 Pgm’ <1 Kimberley Pgm’ < 1 Pgm' <1 Dodson Pgm ° <1 Pgm ' <1 K10 Pgm° <1 Pgm ' <1 MP6 Pgm ’ <1 Pgm ‘ <1 Y. pseudotuberculosis P81 Pst‘ 14.5 Pst' 1.0 l Pst‘ 18.0 Pst' 1.3 MD67 Pst‘ 7.5 Pst' <1 Hale Pst‘ 8.0 Pst' <1 Galligue Pst‘ 15.0 Pst' 1.4 Parkin Pst‘ 14.0 Pstr 1.0 Y. enterocolitica WA Pst’ 8.4 Pst' 2.7 P76 Pst‘ 9.1 Pst' 1.5 E701 Pst‘ 7.1 Pst' <1 E736 Pst’ 8.3 Pst' 1.0 " Number of viable bacteria per HeLa cell. 25 INFECT. IMMUN. vations). This phenomenon. originally made with Pgm‘ mutants (27). first defined the debilitative effect of exogenous iron on the course of infectious diseases and has now been extended to include a variety of other pathogenic microor- ganisms (48. 52). Initially. the explanation of this efi‘ect was generally attributed to fulfilling a nutritional requirement for the cation which is tightly sequestered in vivo (51. 52). However. it is now established (48) that exogenous iron can also enhance virulence by interfering with a variety of nonspecific mechanisms of host defense including blocking synthesis of transferrin (34). preventing chemotaxis of pro- fessional phagocytes (49). and inhibiting intracellular killing by both oxidative (30. 40) and nonoxidative (19) processes. Accordingly. correct assessment of the process by which injected iron reverses the lesions in virulence present in Pgm‘. Pst‘. and Pstr isolates may be requisite to defining the nature of the virulence factors lost by these mutations. Injected iron was assumed to enhance the virulence of Pst‘ mutants of Y. pestis by inhibiting nonspecific mecha- nisms of host defense (7). Evidence favoring this possibility was the discovery that Pst‘ mutants lacked coagulase and fibrinolysin and thus. after typical infection of peripheral tissues via fieabite. would be unable to initiate systemic infection (9. 11). Furthermore. it seems unlikely that the small 6-megadalton plasmid. known to encode pesticin. coagulase. fibrinolysin. and putative immunity plus mainte- nance functions. could also mediate iron transport activities. Accordingly. we suspect that the Pst’ determinant serves to enhance virulence by enabling the organisms to disseminate rapidly from peripheral foci of infection to favored niches within fixed macrophages of lymphatics and intemai organs. This capability would enable the organisms to avoid lethal encounters with professional phagocytes other than macro- phages. In contrast. the Pgm+ factor was predicted to increase virulence by promoting acquisition of iron in vivo (7. 26. 27). Nevertheless. the first direct evidence linking the Pgm+ determinant with transport of this cation is the obser- vation reported here that typical Pgm’ cells do indeed exhibit an increased nutritional requirement for iron. This finding was not anticipated because a different medium. rendered iron deficient by the same process. did not reveal a difference in growth between Pgm’ and Pgm‘ organisms (37). That medium. however. was prepared by avoiding incorporation of organic compounds capable of chelating significant Fe“. In contrast. the medium used in this study contained high levels of a number of biologically significant chelators. including 0.01 M citric acid (53). The presence of these ligands may account for the inability of Pgm~ mutants to multiply after removal of extractable Fe" since over 100 uM FeClz was required for full-scale growth in the initial version (22) of the medium used. An unexpected observation was that iron-deprived Pgm‘ organisms underwent conversion to osmotically stable spheroplasts. Further study will be required to determine whether this phenomenon reflects the enzymatic activity of pesticin. If so. the dramatic loss of viability of Pgm’ mutants in iron-deficient medium may represent an inability to main. tain immunity to the bacteriocin in the absence of iron. Alternatively. normally compartmentalized pesticin may be released after death in iron-deficient medium thereby pro- moting formation of spheroplasts. Evidence supporting this second possibility was the finding that the Pst' determinant was not required for lethality of all Pgm‘ mutants in iron- deficient medium. Indeed. the observation that three Pgm‘ isolates and suppressor mutants of Pgm‘ cells of strain Kuma grew normally in iron-deficient medium at 37°C also VOL. 55. 1987 indicates that the Pgm’ determinant is not essential for transport of the cation in this environment. We are presently investigating the nature of this suppressor mutation and intend to determine whether it restores virulence of Pgm‘ organisms via peripheral routes of injection. Attempts to demonstrate a similar requirement for iron in Pstr mutants of Y. pseudotuberculosis and Y. enterocolitica were not successful. Accordingly. a comparative study was initiated to identify other deficiencies that might account for avirulence in these isolates. No significant differences were noted except in the ability to penetrate nonprofessional phagocytes as typified by HeLa cells. This capability was previously defined in Y. pseudotuberculosis (6) and Y. en- rerocoli'n‘ca (13. 32. 45) and probably reflects that also reported for invasion of other types of nonprofessional phagocytes (25. 38). Further study will be required to determine whether this activity is identical to that trans- ferred to E. coli K-12 as a single genetic locus from a pesticin-insensitive strain of Y. pseudotuberculosis (25). Similar penetration of host cells was not observed with Y. pestis. suggesting that this capability is only required by those yersiniae normally transmitted via the oral route of infection. Nevertheless. it is difficult to reconcile that selec- tion for Pstr in Y. pestis results in recovery of Pgm' mutants. whereas similar selection with Pst‘ isolates of Y. pseudom- berculosis or Y. enterocolitica permitted recovery of mu- tants unable to invade HeLa cells. Before this finding. we suspected that the pesticin receptor per se served as the hemin- and Congo red-binding component unique to Pgm* cells of Y. pestis. This notion became untenable with the discovery that mutational loss of the pesticin receptor in the Other yersiniae resulted in inability to invade HeLa cells. However. a strong possibility remains that the structural gene for the pesticin receptor is linked to genes encoding host-cell invasiveness in Y. pseudotuberculosis and Y. en- terocolitica and linked to distinct genes in Y. pestis that mediate storage of iron. Development of methods permitting precise genetic analysis of chromosomal functions may be necessary to resolve this relationship. The ability of Y. pseudotuberculosis and Y. enterocolitica to invade host cells would. like expression of the Pst+ factor. clearly provide protection against a variety of nonspecific mechanisims of host defense including destruction by pro- fessional phagocytes other than macrophages. However. this activity could also facilitate accumulation of iron as does the Pgm' determinant. Evidence supporting this possibility is tenuous and dependent on the observations that other facultative intracellular parasites including neisseriae (41), listeriae (14). legionellae (39). and salmonellae (4) also utilize siderophore-independent mechanisms of iron transport in vivo possibly analogous to that first described for yersiniae (37). As initially suggested for legionellae (39). these findings are consistent with the hypothesis that available iron is extremely scarce in extracellular spaces but easily obtained in intracellular environments. If this assumption is correct. cells of Y. pestis may obtain and store needed iron via use of the Pgm’ factor. whereas the other yersiniae obtain the cation from intracellular reservoirs. In any event. we previ- ously established that Pstr mutants of Y. pseudotuberculosis and Y. enterocolitica were of reduced virulence (47) and have now shown that this effect reflects the inability to invade nonprofessional phagocytes. ACKNOWLEDGMENTS This work was supported by Public Health Service grant Al 13590 from the National Institutes of Health. 26 PESTICIN RESISTANCE IN YERSINIAE 577 Preliminary experiments concerned with formation of osmotically stable spheroplasts were performed by Robert D. Perry and Susan C. Straley. The excellent technical assistance of Janet M. Fowler is gratefully acknowledged. LITERATURE CITED 1. Beesley, E. D.. R. R. Brubaker. W. A. Janssen. and M. J. Surgalla. 1967. Pesticins. III. Expression of coagulase and mechanism of fibrinolysis. J. Bacteriol. 94:19-26. 2. Ben-Curious, R.. and I. Hertman. 1958. Bacteriocin—like material produced by Pasteurella pestis. J. Gen. Microbiol. 19:289-297. 3. Ben-Gurion. R.. and A. Shaflerman. 1981. Essential virulence determinants of different Yersinia species are carried on a common plasmid. Plasmid 5:183-187. 4. Benjamin. W. 11.. Jr.. C. L. Turnbough. Jr.. B. S. Pusey. and D. E. BriIeS. I985. The ability of Salmonella valirmurrum to produce the siderophore enterobactin is not a virulence factor in mouse typhoid. Infect. Immun. 50:392-397. 5. Berkholl. H. A.. and A. C. Venlal. 1985. Congo red medium to distinguish between invasive and non-invasive Escherichia coli pathogenic for poultry. Avian Dis. 30:117-121. 6. Bovallius. A.. and G. Nilsson. 1975. Ingestion and survival of Yersinia pseudotuberculosis in HeLa cells. Can. J. Microbiol. 21: 1977-2007. 7. Brubaker. R. R. 1979. Expression of virulence in yersiniae. p. 168-171. In D. Schlessinger (ed.). Microbiology—1979. Ameri- can Society for Microbiology. Washington. DC. 8. Brubaker, R. R. 1970. Mutation rate to nonpigmentation in Pasteurella pestis. J. Bacteriol. 98:1404-1406. 9. Brubaker. R. R.. E. D. Beesley, and M. .l. Surgalla. 1965. Pasteurella pestis: role of pesticin l and iron in experimental plaque. Science 149:422—424. 10. Brubaker. R. R.. and M. J. Surgalla. Pesticins. I. Pesticin- bacterium interrelationships. and environmental factors influ- encing activity. J. Bacteriol. 82:940-949. 11. Brubaker, R. R.. M. J. Surgalla. and E. D. Beesley. 1965. Pesticinogeny and bacterial virulence. Zentralbl. Bakteriol. Parasitenkd. Infelttionskr. Hyg. Abt. 1 Orig. 196:302-315. 12. Burrows. T. W.. and G. A. Bacon. 1960. V and W antigens in strains of Pasteurella pseudotuberculosis. Br. J. Exp. Pathol. 39:278-291. 13. Chambers. C. E.. S. L. Stoekman. and D. W. Niecel. 1985. Thermoregulated expression of a cloned Congo red binding activity gene of Shigella flexneri'. FEMS Microbiol. Lett. 28: 281-286. 14. Cowart. R. 5.. and B. 6. Foster. 1985. Differential effects of iron on the growth of Listeria munocyrogenes: minimum require- ments and mechanisms of acquisition. J. Infect. Dis. 151: 721-730. 15. Devenish. J. A.. and D. A. Schiemann. 1981. HeLa cell infection by Yersinia enterocolitica: evidence for lack of intracellular multiplication and development of a new procedure for quanti- tative expression of infectivity. Infect. Immun. 32:48—55. 16. Fer-her. D. M.. and R. R. Brubaker. I979. Mode of action of pesticin: N-acetylglucosaminidase activity. .1. Bacteriol. 139: 495-501. 17. Ferber, D. M.. and R. R. Brubaker. 1981. Plasmids in Yersinia pestis. Infect. Immun. 31:839-841. 18. l-‘erber, D. M.. J. M. Fowler. and R. R. Brubaker. 1981. Mutations to tolerance and resistance to pesticin and colicins in Escherichia coli ¢- J. Bacteriol. 146:506—511. 19. Gladstone, G. P.. and E. Walton. 1971. The effect of iron and haematin on the killing of staphylococci by rabbit polymorphs. Br. J. Exp. Pathol. 152:452-464. 20. Hall, P. 1., and R. R. Brubaker. I978. Pesticin-dependent generation of osmotically stable spheroplast-like structures. .1. Bacterial. 136:786-789. 21. Hertmen. 1., and R. Ben-Gurion. 1959. A study of pesticin biosynthesis. .|. Gen. Microbiol. 21:135-143. 22. Higuchi. K.. and C. E. Carlin. 1957. Studies on the nutrition and physiology of Pasteurella pestis. l. A casein hydrolyzate me- dium for the growth of Pasteurella pestis. J. Bacteriol. 73:122-129. 578 23. 24. 25. 31. 32. 33. 34. 35. 37. 38. SIKKEMA AND BRUBAKER Ha. P. C.. and R. R. Brubaker. 1974. Characterization of pesticin: separation of antibacterial activities. J. Biol. Chem. 249:4749—4753. Hu. P. C.. G. C. H. Yang, and R. R. Brubaker. 1972. Specificity. induction. and absorption of pesticin. 1. Bacterial. 112:212-219. Isberg, R. R.. and S. Falkow. 1985. A single genetic locus encoded by Yersinia pseudotuberculosis permits invasion of cultured animal cells by Escherichia coli K-l2. Nature (London) 317:262-264. . Jackson, S.. and ‘1‘. W. Burrows. 1956. The pigmentation of Pasteurcllu pestis on a defined medium containing haemin. Br. J. Exp. Pathol. 37:570-576. . Jackson. 8.. and T. W. Burrows. 1956. The virulence enhancing effect of iron on non-pigmented mutants of virulent strains of Pasteurclla pestis. Br. J. Exp. Pathol. 37:577-583. . Jackson. 8., and B. C. Morris. 1961. Enhancement of growth of Pasteurella pestis and other bacteria in serum by the addition of iron. Br. J. Exp. Pathol. 37:570-576. . Janssen. W. A.. and M. .l. Surgalla. 1969. Plague bacillus: survival within hast phagocytes. Science 163:950—952. . Kaplan. S. S.. P. G. Quie. and R. E. Basford. 1975. Efl'ect ofiron on leukocyte function: inactivation of H30; by iron. Infect. Immun. 12:303-308. Kol’tsava. E. C.. Y. C. Suchkov. and S. A. Legedeva. 1973. Transmission ofa bacteriocinogenic factor in Pasteurella pestis. Sov. Genet. 7:507-510. Lee, W. H.. P. P. McGrath. P. H. Carter. and E. L. Eide. 1977. The ability of some Yersinia enterocolitica strains to invade HeLa cells. Can. J. Microbiol. 23:1714—1722. Maurelli. A. T.. B. Blackman. and‘R. C. Curtfss III. 1984. Loss of pigmentation in Shi'gella flexncri 2a is correlated with loss of virulence and virulence-associated plasmid. Infect. Immun. 43:397-401. McFarlane. H.. M. Okubadejo. and S. Reddy. 1972. Transferrin and Staphylococcus uureus in kwashiokor. Am. J. Clin. Pathol. 57:587-591. Payne. S. M.. and R. A. Finklestein. 1977. Detection and difl'erentiation of iron-responsive avirulent mutants an Congo red agar. Infect. Immun. 18:94—98. . Perry. R. D., and R. R. Brubaker. 1983. Vwa’ phenotype of Yersinia enterocolitica. Infect. Immun. 40:166-171. Perry. R. D.. and R. R. Brubaker. 1979. Accumulation of iron by yersiniae. Infect. Immun. 137:1290-1298. Portnoy. D. A.. S. I... Moseley. and S. Falkow. 1981. Character- 39. 41. 42. 43. 45. 47. 49. 50. 51. 52. 53. 27 INFECT. IMMUN. 'ization of plasmids and plasmid-associated determinants of Yersinia enterocolitica pathogenesis. Infect. Immun. 31:775- 782. Reeves. M. W.. L. Plne. J. B. Neilands. and A. Balows. 1983. Absence of siderophore activity in Legiunellu species grown in iron-deficient media. .1. Bacterial. 154:324-329. . Schultz. 1.. and S. Rasenthal. 1959. Iron. ll. Inactivation of myeloperaxidase. J. Biol. Chem. 234:2486—2490. Simonson. C.. D. Brener, and I. W. DeVae. 1982. Expression of a high-affinity mechanism for acquisition of transferrin iron by Neisseria meningitidis. Infect. Immun. 36:107-113. Straley. S. C.. and R. R. Brubaker. 1982. Localization in Yersinia pestis of peptides associated with virulence. Infect. Immun. 36:129-135. Surgalla. M. J.. and E. D. Beesley. 1969. Congo-red-agar plating medium for detecting pigmentation in Pusleurellu pestis. Appl. Microbiol. 18:834-837. . Une. '1‘. 1977. Studies on the pathogenicity of Yersinia emera- coh‘u’ca. 1. Experimental infection in rabbits. Microbial. Immu- nol. 21:349-363. Une. T. 1977. Studies on the pathogenicity of Yersinia emera- cuh‘ri'ca. II. Interaction with cultured cells in vim). Microbiol. Immunol. 21:365-377. . Une. T. 1977. Studies on the pathogenicity of Yersinia entero- colillcu. III. Comparative studies between Y. enterocolitica and Y. pseudotuberculosis. Microbiol. Immunol. 21:505-516. Une. T.. and R. R. Brubaker. 1984. In viva comparison of avirulent Vwa‘ and Pgm" or Pst' phenOtypes of yersiniae. Infect. Immun. 43:895-900. . van Asbeck, B. S., and .I. Verhoef. 1983. Iron and host defense. Eur. J. Clin. Microbiol. 2:6-10. Ward. P. A.. P. Goldschmidt, and N. D. Greene. 1975. Suppres- sive effects of metal salts on leukocyte and fibroblast function. RES 1. Reticuloendothel. Soc. 18:313-321. Waring. W. S., and C. II. Werkman. 1942. Growth of bacteria in an iron-free medium. Arch. Biochem. 1:303-310. Weinberg. E. D. 1974. Iron and susceptibility to infectious disease. Science 184:952-956. Weinberg, E. D. 1978. Iron and infection. Microbiol. Rev. 42:45—66. Zahorchak. R. J.. and R. R. Brubaker. 1982. Effect of exogenous nucleotides on Ca” dependence and V antigen synthesis in Yersinia pestis. Infect. Immun. 38:953-959. 28 CHAPTER II Outer membrane peptides of yersiniae and ESCheriChia coli associated with acquisition of iron and sensitivity to pesticin by Daniel J. Sikkema and Robert R. Brubaker 29 ABSTRACT Iron-stress outer membrane peptides required for transport of Fe3+ in Enterobacteriaceae are established receptors of group B colicins. Pesticin is an analogous bacteriocin produced by Yersinia Estis known to inhibit certain strains of Escherichia coli, Yersinia enterocolitica, Yersinia pseudotuberculosis, and nonpesticinogenic mutants of l. Estis capable of absorbing exogenous pigments including hemin and Congo red (Pgm+). Avirulence of pesticin-resistant (Pstr) yersiniae was previously correlated with inability to invade non- professional phagocytes (X. enterocolitica and f_x'_. pseudotuberculosis) or rmtation to Pgm- (X. Estis). In this study, Pstr mutants of g. _c_9_l_i_ p5, unlike isogenic f_ep or Log]; isolates, grew typically in iron-deficient medium and lacked basic outer membrane peptides distinct from iron-stress receptors of group B colicins (_f_ep, _f_h_u;A_, and _c_i_r_). Similar basic structures were also expressed by wild type but not Pst l. enterocolitica WA and l. pseudotuberculosis PBl; these peptides were not observed in outer membranes of Pgm+ or Pgm- X. pestis KIM. Outer membranes of wild type and Pstr l. enterocolitica shared at least nine iron-repressible peptides (Irps) . Similar results were obtained for X. pseudotuberculosis except that expression of a structure corre- sponding to the host cell invasin of this species was repressed by iron and lost upon mutation to Pstr. At least 10 Irps were detected in outer membranes of X. pestis, the major one being the pesticin 30 receptor; it accounted for 10% of the protein in outer-membranes isolated from iron-starved cells of Pgm? X, pestis KIM. Approximately half of the Irps were absent in comparable preparations of Pgmf mutants . 31 INTROIIJCI‘ION The ability of cells of Yersinia pestis, the causative agent of bubonic plague, to absorb exogenous planar small molecules (e.g., hemin, crystal violet, and Congo red) at 26°C and thus grow as pigmented colonies (Pgm+) serves as an important determinant of virulence (33,34,57). Pgm“ nutants exhibit typical lethality in mice by intravenous (50% lethal dose ’\:102 organism) (58) but not by peripheral routes of infection (50% lethal dose >107 organisms) unless serum transferrin is saturated by injection of iron (34). In addition, Pgm- organisms are unable to store hemin (a nutritional source of iron) (44) at room temperature and cannot maintain sustained growth at 37°C in iron-deficient medium (49). Neither Pgm+ (or Pgm—) isolates of 1. Estis nor wild type but phenotypically nonpigmented Yersinia pseudotuberculosis and Yersinia enterocolitica produced detectable siderophores during growth in iron-deficient medium (44,56) although this environment induced typical expression of outer membrane iron- repressible peptides (Irps) (9). One structure composing 10% of the protein in the outer membrane of iron—starved cells of Pgm+ 1. Egg is the pesticin receptor. This structure became separable from the pigmentation phenotype by selecting pesticin resistant mutants which remain pigmented on Congo red-pesticin agar plates (57,58). The additional Irps as well as the pesticin receptor may mediate the 32 siderophore-independent cell-bound high-affinity process of Fe 3+ transport previously defined in yersiniae (44). Bacteriocins of ESCherichia coli, termed colicins (21), are separated into groups A and B depending upon the nature of their corresponding outer membrane absorption sites (13,14). SuCh receptors of group B colicins are typically involved in iron metabolism. For example, mutational loss of the structure required for uptake of the endogenous phenolate siderophore enterochelin (£29) or that for exogenous hydroxamate siderophores (ghgé) promotes resistance to colicins B or D and colicin M, respectively (13). Similarly, gig mutants which lack a third iron-stress outer membrane receptor of unknown physiological significance (53) are unable to absorb group B colicins Ia, Ib, Q, 81, and V (13). This phenouenon of resistance is distinct from that of tolerance where insensitivity reflects loss of functions required for penetration of bacteriocins to internal targets or modification of snob targets (36). A salient example of tolerance in g, 99;; is mutation to Egg§_resulting in loss of an inner membrane function (46,60) required for assindlation of all group B colicins (13) and acquisition of iron by siderophore-dependent (13,37) and other high—affinity mechanisms of transport (22). Pesticin (2) is a bacteriocin produced by wild type 1. m (Pst+) capable of inhibiting growth of serotype IA and IB strains of 1} pseudotuberculosis (8), serotype 0:4,32, 0:21, and 0:8 strains of X, enterocolitica (26,30,45), and certain isolates of g, 9911 (6,30,52) including the universal colicin indicator strain d'of Gratia (23). 'nme nature of the pesticin receptor in these organisms is not yet resolved. However, several findings outlined below suggest that this structure, 33 like those for group B colicins, my be required for assimilation of iron. First, lethality of pesticin was prevented by exogenous Fe3+ or hemin suggesting that its receptor is repressed by the cation (6,30). Second, although resistance to pesticin (Pstr) was not shared by £6.32, _f_r_1_uA_, or gig; mutants of _E_:. _c_o_l_i_ {6, top; isolates of this organism were also tolerant to the bacteriocin (19). Third, mutation to Pstr in Pst- isolates of l. Estis (which lack inmmity to the bacteriocin) (15) resulted in avirulence due to high frequency conversion to Pgm- (5) . Finally, Pstr mutants of X. pestis (4) which retain pignentation are unable to sustain growth in highly iron—deficient medium and thus resemble Pgm- cells (this study). Pstr mutants of _Y_. pseudotuberculosis and X. enterocolitica were subsequently found to exhibit reduced lethality by peripheral routes of injection (58). However, avirulence in this case was not correlated with a detectable lesion in iron metabolism but rather reflected inability to invade nonprofessional phagocytes (49) , a function promoted by defined outer membrane host cell invasins (31 ,32,50). The purpose of this study was to compare the pesticin receptor of E. coli and yersiniae and to define the nature of the diverse Pstr phenotypes expressed by these species. Results demonstrated that iron-stress phenotypes of yersiniae, especially 1. Estis, are more complex than that of g. c_oli, a peptide corresponding to the host cell invasin of l. pseudotuberculosis is lost upon mutation to Pstr and is regulated by iron, and that about half of the approximate 10 iron- stress peptides of 1. pestis are eliminated upon imitation to Pgm: one of which is the pesticin receptor. 34 MRTERIALS AND METHODS Bacteria. All yersiniae used in these determinations were avirulent due to mutational loss of the m45~negadalton plasmid that mediates the lowecalcium response (18). Type strains employed were 1. pestis KIM (variety mediaevalis) (4), X. pseudotuberculosis PBl (serogroup IB) (8), and X} enterocolitica WA (serogroup 0:8) (10). The universal colicin-indicator strain E, ggli_d'(21,23) and its bacteriocin-resistant and tolerant mutants have been described (19). An auxotroph of g. c_o_l_i_ «1 blocked in the synthesis of hemin was induced by exposure to ultraaviolet light; its population was enriched by growth in the presence of 6-amdnolevulinic acid followed by recycling without this compound in medium containing penicillin. Clinical isolates of g, ggli_were primarily of enteroinvasive phenotypes. Pstr mtants of X. pseudotuberculosis and X. enterocolilca and the Pgm- nutant of I} p§§t1§_were the same as those previously described (49, 58). Pstr, Pgmf mutants of X} p§§§i§_were isolated on Congo red- pesticin agar (49,57,58). Cultivation. Cells of g. _cgl_i_ d were routinely grown in appropriately supplemented (17) morpholinOpropane sulfonic acid— buffered chemically defined medium (41). All yersiniae were cultivated in the synthetic medium of Higuchi et a1. (28) as modified by Zahorchak and Brubaker (61). Iron was omitted during preparation of these media when used for expression of iron-stress functions and, after 35 TABLE 1. Pr0perties of bacterial strains. Pigmen- Produc- Sensi- Biotypeb Species Strain tation tion of tivity to or sero- pesticin pesticina type 1. pestis KIM + + I M KIM + - S M KIM + - R M KIM -— + I M KIM - — R M X. pseudotuberculosis PBl - - S I PBl - - R I X. enterocolitica WA - - S 0:8 WA - - R 0:8 E. coli g5 - - S - _ _ R _ a. I, iumune; S, sensitive; R, resistant. b. Biotype is mediaevalis (M). 36 preparation, such media were further extracted with 8- hydroxyquinoline (58) leaving a level of < 0.3 uM iron as determined by flame absorption spectroscopy. Bacteria were retrieved from stock cultures, inoculated, and subcultured as previously described (49); final transfers where the organisms were grown for use in electro- pherograms were always performed at 37°C. 35S—methionine (20 uCi/ml) Cells of I, pestis were labelled with and grown to constant specific activity (at least 10 generations). Densitometric analysis of autoradiograms and silver stained gels allowed the determination of the percentage of total outer-membrane protein synthesis accounted for by individual peptides. Preparation of outer membranes. Procedures used for fractionation of enteric bacteria are established (43,47) and were described in detail for use with yersiniae (54). Briefly, the method involves conversion of organisms to spheroplasts with lysozyme and EDTA followed by disruption by sonication. Particulate material was collected by centrifugation and then separated into inner and outer membranes by iSOpycnic sucrose-gradient centrifugation. Contamination of outer membrane preparations by inner membranes is about 3% (54). Electrophoresis. The method of O'Farrell (42) was used for preparation of electropherograms; details of this process as used with cytoplasm, inner membranes, and outer membranes of I, pestis grown in excess iron have been reported (54,55). As used in this study, the procedure was especially effective in resolving the high molecular weight peptides of low isoelectric point typically induced in iron- deficient media. 37 Miscellaneous. Protein in samples used for electrophoresis was determined by the method of Lowry et al. (38) and peptides in electropherograms were visualized by silver staining (40). Pesticin was prepared as previously described (29,30) and assayed against cells of X, pseudotuberculosis PBl on solid medium containing EDTA in excess 2+ ( Ca 7). Colicins were induced, isolated, and, when necessary, separated as previously reported (19). 38 RESULTS Pstr nutants of _E_. coli. Initial work involved defining the putative pesticin receptor of _E_. ggl_i_. ElectrOpherograms were prepared by using outer membranes of strain a! grown with excess Fe3+ (Fig. 1A) and in iron—deficient uedium (Fig. 18). At least three iron-stress peptides were induced in the latter which were absent in similar pre- parations of isogenic fep (Fig. 2A), fhuA (Fig. 2C), or cir (Fig. 2D) mutants. Loss of an analogous well—defined structure was not detected in the Pstr mutant although two peptides of more basic isoelectric point appearing as streaks were absent in this isolate (Fig. 2B). These structures were not fully repressed by excess Fe 3+ in the parent (Fig. 13) although their production was pronounced under conditions of extrelre iron privation as in the £92 mutant grown in extracted uedium (Fig. 2A). To determine if these peptides are required for accumulation of iron, growth of the parent in iron-deficient and iron-supplemented medium (Fig. 3A) was compared to that of isogenic Pstr (Fig. BB), f_ep (Fig. 3C), and £o_n_B_ (Fig. 3D) mutants. Maximum optical densities of all cultures were similar after growth with excess FeCl3 (20 uM). These values were reduced by about half for wild type and Pstr organisms in unsupplemented medium and, as expected, were markedly decreased in cultures of _fgp and 39913 nutants. Intermediate maximum Optical densities were obtained upon addition of hemin (20 pM) to 39 .1 t" .v 2 ‘t i \ 1 - *3 w , t - ‘I'n .' h .3 . I) ' § .. i l v» u'r» p. r ; ii \' ’ .“. d r" t I1 It} I: a . I ‘. £9 t, v;. . . I '_' r l ‘v . ’ pr , | .j.‘ .I‘ -. 1' . I h . ‘I fi' ‘ $4 -95; ,- ' ~5er £933..."- '1'. "‘1 I. - V L . .- , ’5’{.ae&_¢f'-TW$&.$ .;: ”.«.’-‘-kl\hb"‘.~'€"5- “8,!- 7 r, 9 i 6 i v .6 .__40 Figure 1. Electropherograms (two-dimensional) of outer membranes of Escherichia coli strain ¢ grown at 37°C (A) in the presence of excess FeCl and (B) in the absence of FeCl . Arrows indicate the absence of known iron-repressible3peptides. 4:---" 505 Figure 2 . 40 Electropherograms of outer membranes of (A)f eB,p (B) Pstr , (C) Fh__u_A_, and (D) c__i__r nutants of Escherichia coli strain :6 grown in iron-deficient medium. Arrows indicate loss of corresponding outer-membrane receptor . OPTICAL DENSITY .0 ad. 41 .4 O I WTTjTfi fl I l 1111] 1 441444] 1 ii 111r l 0 qt- “r- P 'jr- p 1r- "F' qr- '4!- 1P qr- J JJJJJJ 1 11111] 1 A 1.1 llJl 7 1 jfiijII j 11111] p- l #JJJJJII J JJJJJI 0 Figure 3 . 1b 2b 0 1b 2b 0 HOURS 1b-éo 0 1b éo Growth of (A) wild type, (B) Pstr, (C) fepB, and (D) tonB cells of Escherichia coli strain ,6 in iron-deficient medium (0) or the sane nedium containing‘added hemin (a) or excess F8013 (O)- Absorbance was measured at 620 nm. 42 iron—deficient cultures of wild type, Pstr, and gap isolates. This source of iron did not promote significant growth of Eggg mutants. Results obtained in similar experiments with double Pstr, gap or Pstr, £222 mutants were identical to those Shown in Fig. 3 for single §§p_and Egg§_isolates, respectively (not illustrated). These similar patterns of growth.obtained for wild type and Pstr mutants indicate that the basic peptides unique to the former are not required for accumulation of Fe3+. However, the ability of hemin to promote growth of wild type, Pstr, and §§p_isolates was not antici- pated. To determine if internalized hemin per se accounted for this increase in optical density, we selected an auxotroph of the pesticin-sensitive parent blocked in synthesis of the hemin precursor 6-aminolevulinic acid. This hem mutant grew normally from low inocula (optical density of «0.05) in defined uedium containing 0.1 nM 6-aminolevulinic acid (maximum Optical density of m2.5) but failed to multiply in the presence of 0.1 to 2.5 mM protOporphyrin IXIor‘hemdn (maximum optical density of 0.1) unless 5-aminolevulinic acid was included. This Observation indicated that hemin was not internalized at a sufficiently rapid rate to permit its use as a source of iron. Accordingly, the ability of hemin to enhance growth in iron-deficient medium may reflect extracellular removal of covalently bound iron. To further resolve the nature of the mutation to Pstr in g, 921;, a series of 40 clinical isolates was screened for sensitivity to pesticin. Growth of half Of these was inhibited by high concentrations of the bacteriocin (105‘Uflml). ‘nhis frequency was significantly greater than that observed for normal human isolates (none of 100 43 strains, 19) suggesting that the clinical isolates might possess a unique receptor for the bacteriocin. Further study showed that the clinical isolates were typically inhibited by colicin M but insensitive to 13 additional colicins of both group A and B. In contrast, the normal strains were generally sensitive to most of these activities. Those clinical strains that were sensitive to pesticin yielded Pstr mutants capable of growth in the presence of high concentrations of the bacteriocin (105 U/ml). These mutants were also insensitive to colicin M and capable of growth in iron-deficient medium comparable to that of wild type. In addition, patterns of outer membrane iron-stress peptides observed in electropherograms of these Pstr mutants were identical to those of the parents. Further study Showed that the Pstr mutant of E. _cc_>_l_i_ d, initially selected for resistance to 2 x 104 U of pesticin per ml, was also inhibited by the higher concentration used in this study (105 U/mu). These findings indicate that sensitivity of the clinical isolates and the Pstr mutant of E. c_oli d to high concentra- tions of pesticin is a nonspecific phenomenon distinct from that accounting for sensitivity of wild type _E. _c_o_l_i d and yersiniae to low levels of the bacteriocin. Pstr mutants of X, enterocolitica. Electropherograms were produced from outer membranes of X} enterocolitica'WA grown with excess Fe3+ (Fig. 4A) and in iron—deficient medium (Fig. 4B). The response of this organism during iron-privation was mre complex than that Of _E. 99;; as judged by induction of at least 9 major iron—stress peptides. All of these structures were present in outer nemhranes of isogenic Pstr organisms cultivated under identical conditions (Fig. 4C). However, this untant lacked a set of basic peptides similar to ‘—-- 505 Figure 4 . 44 Electropherograms of outer membranes of Yersinia enterocolitica strain WA grown (A) with excess FeCl3, (B) in iron-deficient medium, and (C) outer mahranes of an isogenic Pstr mutant grown in iron-deficient medium. Peptides within boxed region represent structures associated with sensitivity to pesticin. Arrows indicate major iron-stress peptides. 45 those previously associated with expression of sensitivity of E. 99;; d to pesticin (Fig. 2B). These structures were also present, but at reduced concentration, in the parent grown with excess Fe3+ (Fig. 4A). Pstr mutants Of E. pseudotuberculosis. Outer uembranes from cells of E. pseudotuberculosis PBl grown with excess Fe3+ (Fig. 5A) and in iron-deficient medium (Fig. 5B) were similarly compared in electro- pherograms. Approximately 9 major iron stress peptides, evidently analogous to those of E. enterocolitica were induced in the latter which also contained basic structures similar to those Observed in wild type E. 9&1; d (Fig. 2B) and _Y_. enterocolitica (Fig. 4C). In addition, this organism expressed a distinct iron-stress outer nembrane peptide Of about 100,000 daltons (Fig. SB) equivalent to that reported for the host cell invasin of E. pseudotuberculosis (32) . Outer membranes of the isogenic Pstr mutant shared all of the major iron-stress peptides except for that corresponding in molecular weight to the host cell invasin (Fig. 5C). The Pstr mutant of E. pseudotubercul.o_s_i_s (Fig. 5C), like the Pstr isolates of E. _c9_l_i_ (Fig. ZB) and E. enteroco;itica (Fig. 4C) , also lacked detectable levels of the basic outer membrane peptides previously equated with sensitivity to pesticin. Pstr nutants of _Y_. pgstis . Electropherograms were prepared of outer membranes of Pst+ , Pgm+ cells of _Y_. Eéé‘flélé KIM grown in excess Fe3+ (Fig. 6A) and in iron-deficient nedium (Fig. GB). Approximately 10 unique peptides were induced in the latter, none of which corre— sponded in position to the putative host cell invasin of E. gendotuberculosis shown in Fig. SB. Electropherograms of outer + - membranes of Pst , Pgm isolate prepared after growth in excess 46 <---SDS Electropherograms of outer membranes Of Yersinia geudotuberculosis strain WA grown (A) with excess FeCl3, (B) in iron-deficient nedium, and (C) outer nembranes of an isogenic Pstr nutant grown in iron—deficient medium. Boxed region encloses structures previously associated with pesticin resistance; circular surround indicates a peptide corresponding to a M.w. of "100,000 daltons analogous to invasin. Arrows indicate major iron—stress peptides. Figure 5 . 46a Figure 6. Electropherograms of outer membranes of Pgmf, Pst+ cells of X, pestis strain KIM grown at 37°C (A) in the presence + . of 50 “M Fe3 and (B) in iron-deficient medium containing 3+ <0.3 pM Fe . See Table 2 for characteristics of iron- repressible peptides A—J. IEF-----h- 8%---.) Figure 6 . 48 TABLE 2 . Iron-repressible outer-membrane peptides . a b Isoelectric Percent: of Irp M.w. pOlnt 0MP Cell type A 67.7 5.45 10 pgm+ B 68.0 5.35 2.0 pgm+ c 68.2 5.56 2.0 pgm+ D 65.1 5.98 1.0 Pgm+ E 40.3 6.63 0.5 Pgm+ F 17.9 5.89 0.5 Pgm+ G 65.8 5.54 0.3 Pgm+ H 80.2 5.20 1.0 Pgm+ s. Pgm- I 75.3 6.65 1.5 Pgm+ s. Pgm- J 34.1 7.10 1.5 Pgm+ s. Pgm- a. Iron-repressible outer membrane peptides b. M.W. = molecular weight in daltons c. OMP = outer membrane peptide synthesis TABLE 3. Iron-inducible outer-membrane peptides. Percent of a b Isoelectric Iip M.w. point 0MPC A 17.9 4.71 14 B 23.4 5.34 4 c 34.2 4.81 4 D 32.8 5.99 2 E 45.1 5.20 7 a. Iron—inducible outer-membrane peptides b. M.w. = molecular weight in daltons c. 0MP = outer nembrane peptide synthesis 49 Fe3+ (Fig. 7A) and in iron-deficient uedium (Fig. 78) illustrated that greater than half of the iron-stress peptides shown for the Pgm+ parent were lost upon selection for resistance to pesticin. Electropherograms of outer membranes of the corresponding Pst—, Pgm+ isolate grown in the presence of excess Fe3+ (Fig. 8A) and in iron—deficient medium (Fig. 8B) demonstrated that an Irp with a molecular weight of 34.1 and an isoelectric point of pH 7.1 is uediated by the 10 Kb pesticin plasmid and may be the pesticin immunity factor. This peptide is also absent in similar preparations of Pst-U Pgm- isolates (Fig. 8C). Structures analogous to the basic peptides equated with sensitivity of E. £11; 2! and the other yersiniae to pesticin were not detected in outer membranes of E. @143. However , electrophero- grams of Pst-, Pgm+, Pstr isolates of 1. Eggs strain KIM confirm that the pesticin receptor is the ma j or iron—repressible outer-membrane peptide (Fig. 9) and that growth in an iron-deficient enviromrent is dependent upon its presence (Fig 10). Characteristics of the pesticin receptor are a molecular weight of 67,700 daltons and an isoelectric point of 5.45. Synthesis of the pesticin receptor is increased 100- fold in iron-extracted nedium (<0.3 UM) when compared to growth in the presence of 50 11M Fe 3+. 49a Figure 7. Electropherograms of outer membranes of Pgm-. Pst+ cells of E. pestis strain KIM grown at 37°C in the (A) presence of 50 pM Fe3 and (B) m lron-def1c1ent medium containing <0.3 11M 1783+. IEF-----~ ".‘i‘fvSDS 4.»: .' . ,- .54.... .‘ . . * .“| ' . . . .V. n - _ . , 5. . _ . l . , . '. -- . u . ' ' . .' p.- - n - . . 1 . ‘ .; , 1| ', . . . . .. - 0.. -. . r ' " . n h " . ' ..: . ‘I ~ .. ,0 9 {J}; ‘I‘ I I J " ~ "I 9' a' \‘- AWN? u.‘ -.-'.n‘ a [I J Figure 7 . Figure 8 . 50a Electropherograms of outer membranes isolated from (A) Pgm+, Pst', (B) Pgm+, Pst’, and (c) pgm", Pst' cells of _Y_'_. Estis strain KIM. Iron-repressible structures associated with pesticinogeny are absent. Membranes comprising (A) were prepared from cells grown in the presence of 50 pM Fe3+; (B) and (C) were grown in iron-extracted medium containing <0.3 “M Fe3+. ‘--SDS M‘- -‘, 1 Figure 8 . 52 IEF—----)-— mew---) . + Flgure 9. Electropherograns of outer membranes of Pgm , Pstr cells Structures associated with of E. estis strain KIM. Irp (F) was not shown absorption of pesticin are absent. Fig. 6 due to its low molecular weight (see Table 2). 1n Figure 10 . 52a Growth comparison of (A) Pgm+, Pst+, (B) Pgm+, Pst’, (C) Pgm+, Pst', Pstr, (D) Pgm", Pst+, and (E) Pgm", Pst- cells of E. past—is. strain KIM in chemically iron- deficient medium (0) or the same nedium containing added hemin 50 “M (A) or added Fe3+ 50 11M (0). Absorbence was measured at 620 nm. 53 zo_._.