Protein Sources of Heme for Haemophilus influenzae

Transcription

1 INFECTION AND IMMUNITY, Jan. 1987, p /87/ $02.00/0 Copyright 1987, American Society for Microbiology Vol. 55, No. 1 Protein Sources of Heme for Haemophilus influenzae TERRENCE L. STULL Division of Infectious Disease, Department of Pediatrics, and Department of MicrobiologylImmunology, Medical College of Pennsylvania, Philadelphia, Pennsylvania Received 13 June 1986/Accepted 7 October 1986 Although Haemophilus influenzae requires heme for growth, the source of heme during invasive infections is not known. We cqmpared heme, lactoperoxidase, catalase, cytochrome c, myoglobin, and hemoglobin as sources of heme for growth in defined media. The minimum concentration of heme permitting unrestricted growth of strain Ela, an H. influenzae type b isolate from cerebrospinal fluid, was 0.02,ug/ml. Using molar equivalents of heme as lactoperoxidase, catalase, cytochrome c, myoglobin, and hemoglobin, we determined that myoglobin and hemoglobin permitted unrestricted growth at this concentration. To determine the ability of host defenses to sequester heme from H. influenzae, we used affinity chromatography to purify human haptoglobin and hemopexin, serum proteins which bind hemoglobin and heme. Plate assays revealed that 12 strains of H. influenzae acquired heme from hemoglobin, hemoglobin-haptoglobin, heme-hemopexin, and heme-albumin. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of outer membrane proteins of strain Ela grown in heme-replete and heme-restricted conditions revealed a heme-repressible outer membrane protein with an apparent molecular mass of 38 kilodaltons. These results demonstrated that, unlike Escherichia coli, H. influenzae may acquire heme from hemoglobin-haptoglobin. H. influenzae also may acquire heme from hemopexin and albumin, which have not been previously investigated. The role of outer membrane proteins in the acquisition of heme is not yet clear. Haemophilus influenzae colonizes the mucosa of the upper respiratory tract of 50 to 80% of adults and children (32). The prevalence of upper respiratory tract colonization by H. influenzae type b (Hib) is only 2 to 4% (32); however, Hib is a common cause of bacteremia in young children. Bacteremia due to Hib may subsequently cause sepsis or localized invasive diseases such as meningitis (25, 26). Intravascular replication may play an important role in the early steps of the pathogenesis of invasive disease (21), and intravascular bacterial density may be a determinant of focal invasive disease (16). Heme is an essential nutrient for H. influenzae (9). (Heme in this manuscript indicates ferrous or ferric protoporphyrin IX.) However, heme is hydrophobic and it is contained by heme proteins such as hemoglobin, catalase, and cytochromc c, which function in oxygen transport, detoxification of oxygen derivatives, and electron transport, respectively. In addition, heme is bound by the serum proteins hemopexin (Kd, < 10-12) and albumin (Kd, -5 x 10-7), and hemoglobin binds to haptoglobin with such affinity that the Kd has not been determined (1, 15, 23). Although the role of hemopexin and albumin in host defenses against microbial invasion has not been investigated, binding of hemoglobin to haptoglobin effectively sequesters iron from Escherichia coli (8). The goals of this investigation were to determine which heme-containing proteins may be sources of this essential nutrient for growth of H. influenzae and to determine the ability of hemopexin, albumin, and haptoglobin to serve as host defenses by heme starvation of H. influenzae. MATERIALS AND METHODS Bacteria and culture conditions. All strains required heme and,-nad for growth. Typing was determined by agglutination with Difco (Detroit, Mich.) antisera. Growth curves were performed with strain Ela, a streptomycin-resistant, stably encapsulated cerebrospinal fluid (CSF) isolate (29). Other CSF isolates included C65 from Seattle, Wash., C148 from the Medical College of Pennsylvania, Philadelphia, Pa., 148 and R152 from Charles Prober, Toronto, Ontario, Canada. The unencapsulated strains C120, C132, and C136 were obtained from eye cultures and respiratory tract cultures at St. Christopher's Hospital for Children (Philadelphia, Pa.). Strains R211, R212, R213, and R214 were type b throat isolates from healthy children, as previously described (27). Supplemented brain heart infusion (Difco) consisted of brain heart infusion broth containing heme and P-NAD (each at 10 Rg/ml). Haemophilus defined medium (HDM) was a modification of the medium described by Catlin (4). All of the described reagents were included except EDTA, sodium lactate, glycerol, polyvinyl alcohol, spermine, 2,2'2'-nitrolotriethanol, and Tween 80. The final concentrations of methionine and sodium acetate were 0.01 and 1 mm, respectively. Freshly prepared heme was added separately in the quantities described for each experiment. Plates were prepared from media contaiing 1.5% agar (Difco). After plating onto brain heart infusion agar containing heme and 3-NAD (each at 10 p.g/ml) from a stock maintained in skim milk at -70 C, the plate was incubated overnight at 37 C in 5% CO2 and stored at 4 C. Subsequent cultures were taken from this plate within a week. On the day before each experiment, organisms were depleted of heme by plating once on brain heart infusion agar containing,-nad and lacking heme. After anaerobic incubation at 37 C for 18 h in a GasPak (BBL Microbiology Systems, Cockeysville, Md.), the organisms were suspended from the plate in heme-free HDM at a bacterial density of approximately 109/ml. For growth in broth, this suspension was added to 10 ml of heme-free HDM until a net change of 20 Klett units was achieved (approximately 0.4 ml). The flasks were incubated at 37 C in a shaking water bath at 200 cycles per min. Growth was monitored with a Klett-Summerson colorimeter with a 540-nm filter. Plate assays. To evaluate the availability of heme from protein sources, organisms were suspended in molten hemefree HDM agar at a density of 107 CFU/ml. Amounts of 10 to 100 RI of each heme source, containing 0.1,ug of heme, were

2 VOL. 55, 1987 placed in 6-mm (diameter) wells. For separate comparison, a well in each plate contained an equal quantity of free heme or hemoglobin. To assure that the heme source did not contain substances toxic to H. influenzae, a separate well contained an equal quantity of heme source and free heme. Plates were incubated for 48 h in 5% CO2 at 37 C. Longer incubation did not significantly alter the results. Results were expressed as the mean diameter of growth zones in duplicate assays. Control wells containing HDM, hemopexin, albumin, or haptoglobin showed no growth zones. Reagents. Unless stated otherwise, all reagents were obtained in the purest grade available from Sigma Chemical Co. (St. Louis, Mo.). Whale myoglobin was obtained from Mann Research Laboratories (New York, N.Y.). Bovine lactoperoxidase was dialyzed twice against 100 volumes of distilled water to remove the preservative. Human serum albumin, equine cytochrome c, bovine catalase, human hemoglobin, and whale myoglobin were used without processing. Each heme protein was solubilized in water, filter sterilized, and used within 24 h. Heme content was confirmed by the pyridine-hemochromagen method (20). Equine heme was solubilized at a concentration of 1 mg/ml in 0.02 N sodium hydroxide containing L-histidine (1 mg/ml). Heme saturation of human serum albumin was calculated by assuming 1:2 molar binding and a molecular weight of 60,000 (18). Human serum hemopexin was purified by affinity chromatography as described by Tsutusi and Mueller (31) with modifications. Briefly, 200,ug of heme was dissolved in 15 ml of N,N-dimethylformamide (spectrophotometric grade), and 200 mg of N,N-carbonyldiimidazole was added. The mixture was heated for 15 min at 80 C and then cooled to room temperature. Aminoethylagarose (10 ml) washed with N,Ndimethylformamide was added, and the coupling reaction was allowed to proceed for 18 h. The heme-agarose was washed in decreasing concentrations of N,N-dimethylformamide and, finally, 200 ml of distilled water. After being washed with 200 ml of 25% pyridine, the heme-agarose was washed with distilled water and stored in the dark at 4 C. Serum (5 ml) diluted 20-fold in 0.5 M NaCI-10 mm sodium phosphate (NP buffer), ph 7.5, was delivered to a column containing 2 ml of heme-agarose. After being washed with 200 bed volumes of NP buffer, the hemopexin was eluted with 8 M urea in NP buffer, ph 6.0. After extensive dialysis against deionized, distilled water, the solution was concentrated to 2 ml by vacuum dialysis. The solution was sterilized by brief exposure to UV light; a sample was cultured to assure sterility, and the remainder was stored at -20 C. Hemopexin saturation was calculated by assuming 1:1 molar binding and a molecular weight of 57,000 (22). Hemopexin was incubated with heme for 20 min at room temperature before nondenaturing polyacrylamide gel electrophoresis (PAGE) or testing in bioassays. Hemopexin function was confirmed by altered heme mobility in nondenaturing PAGE, and the purity was 70% or greater as determined by sodium dodecyl sulfate (SDS)-PAGE (Fig. 1). Heme contamination of hemopexin was undetectable by the pyridine hemochromagen method, indicating less than 10% saturation. Heme was added to the hemopexin preparation to achieve 50% saturation of the hemopexin for use in growth experiments. Human serum haptoglobin was purified with the affinity resin described by Delers et al. (6). Haptoglobin was also obtained from Calbiochem-Behring (La Jolla, Calif.). Haptoglobin was incubated with hemoglobin at room temperature for 20 min before the bioassays. As the results with commercially available haptoglobin were similar to those HEME SOURCES FOR H. INFLUENZAE 149 obtained with the purified preparations, the results were pooled. Haptoglobin saturation was calculated by assuming 1:1 molar binding of subunits with a molecular weight of 85,000 (33). Outer membrane proteins. Outer membrane proteins were prepared with Triton X-100 from organisms grown to stationary phase in a replete concentration of heme (10,ug/ml) or a growth-limiting heme concentration (0.005,ug/ml) as previously described, with minor modifications (28). Growth restriction was defined as a slower growth rate and lower final bacterial density compared with growth in replete concentrations of heme. Briefly, organisms from 50 to 150 ml were pelleted by centrifugation at 15,000 x g for 15 min and suspended in 5.0 ml of 0.75 M sucrose-10 mm Tris-acetate, ph 7.8 (T buffer), containing 0.75 M sucrose, 5 mm MgCl2, 100 jig of DNase per ml, 100 p.g of RNase per ml, and 100,ug of lysozyme per ml. Cells were ruptured by sonification six times for 15 s with a salt water bath for cooling. After centrifugation at x g for 15 min, crude membranes in the supernatant were pelleted at 90,000 x g for 1 h. The pellet was suspended in 2 ml of 2% Triton X-100 in T buffer. After 30 min at room temperature the solution was centrifuged at 90,000 x g for 1 h. The pellet was again extracted with 2% Triton X-100 in T buffer. The final pellet was suspended in water and stored at -70 C. Analytical methods. SDS-PAGE was performed as previously described (28). Nondenaturing PAGE was performed by the same method with the following modifications. (i) The A B C.:...~... Hpx HSA FIG. 1. (I) Nondenaturing PAGE of (lane A) purified human serum hemopexin (Hpx), (lane B) human serum, and (lane C) human serum albumin (HSA). Each sample was preincubated with heme for 20 min at room temperature. The location of the heme was detected by staining for peroxidase activity with the chromogenic substrate 3,3',5,5'-tetramethylbenzidine. (II) SDS-PAGE of purified human serum hemopexin detected by silver staining. Molecular masses were determined by coelectrophoresis with phosphorylase b (92 kda), bovine serum albumin (66 kda), ovalbumin (45 kda), and soybean trypsin inhibitor (21 kda) II

3 150 STULL INFECT. IMMUN. (Di 60 * o c 60*_-4 Home 1.00 jig/mi sti// 0-0 Heme 0.02 jig/mi &--a Heme 0.00 jig/mi m 40 0po-C Heme jig/mi 201 -_ Time (hours) FIG. 2. Concentration of heme for unrestricted growth of Hib strain Ela in HDM (*, P < 0.01 compared with 1.0,ug/ml and *,P < compared with no heme). The results are shown as the mean ± the standard error of the mean calculated after logarithmic transformation A He44_ 0-0 me CT(hours reu) a wa o f t* Home+bCytochrome C o of Wy ta. () Catalase+Heme sample was not boiled, and (iii SO Cytochrome C b D by t p o Lactoperoxidasc h sa bf a e0gl b e 3'5' +Lactoperoxidaso reducing agent was omitted from the sample buffer, (ii) the method of Wray et al. (34). Bands containing heme were sample loo wa's not boiled, and (iii) SDS --60 was 1 omitted 2 3 from both 6 detected20- by the presence of peroxidase activity with the sample buffer and the gel buffers. 3,3',5,5'-tetramethylbenzidine as a chromogenic substrate Protein bands were detected by silver staining by the (30). The sensitivity ofthis strain was enhanced by reduction cj- 40L m,bheeoogl A 40 ~~~~~~~Heme 0.00 mg/mli B B Hems 0.00 JAg/ml 0-0 Hemos Mo_ +-shal oob m60--0* Heme 60-hoe ClaeHeme 0,cthheme Time (hours) Time 0-C anssr Le (hours) P 0.0eme compae d Ir-sfaptoglb rwhrsoneo i tanel nhmcnanngn ee eea1002jgmasfehmeor(nla)cthoec withee a 0.2 jlg/l);(pael ) emoghems heehmpxno eolbnhpogoi.terslsaesona as Hmyoglobin (* 0-.2ad *.2coprdwt o heme) or HempaelDn heeheoexno hmglbnhatglbn Them aesumolobarison ash hema en±tesanaderroh h tnaderroh mean mencac cacuaemeasftemoglobinhi +ae Haftoglobain transformation. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~ 4

4 VOL. 55, 1987 FIG. 4. Growth response of Hib strain Ela (107 CFU/ml) to heme-containing compounds in HDM-agar. Each well contained 0.1,ug of heme. A, free heme; B, heme-hemopexin; C, heme-human serum albumin; D, hemoglobin; E, hemoglobin-haptoglobin. of the gel size with 50% methanol before incubation with 3,3',5,5'-tetramethylbenzidine as described by Palumbo and Tecce (19). Heme was quantitated by the pyridine hemochromagen method with equine heme as the standard (20). Protein was quantitated by the method of Lowry et al. (17) with bovine serum albumin as the standard. Statistical methods. Comparisons of bacterial densities were performed with Student's t test after logarithmic transformation. RESULTS Heme requirement of Hib. Although heme is frequently provided in excess for growth of Hib in vitro (21), the growth rate and final density of strain Ela were similar in HDM containing 0.02 or 1,ug of heme per ml (Fig. 2). Further reduction in the heme concentration to 0.005,ug/ml reduced the growth rate and final bacterial density. Heme at 0.02,ug/ml was used in subsequent experiments to compare the growth rates allowed by protein sources of heme. Acquisition of heme from heme proteins. To determine the potential protein sources of heme for strain Ela, proteins representative of the physiologic functions of heme proteins were selected. The growth of strain Ela in HDM broth with defined sources of heme is shown in Fig. 3. Growth with cytochrome c as the sole source of heme was not significantly different from that in the control flask containing no heme. Lactoperoxidase and catalase, which function to detoxify certain oxygen-containing compounds, were also poor sources of heme for strain Ela. Catalase as the sole source of heme allowed growth after 2 h, in contrast to the flask lacking heme (P < 0.02); however, growth after 2 h was not as great as that in the flask containing free heme (P < 0.02). To HEME SOURCES FOR H. INFLUENZAE 151 determine the potential of heme proteins which transport oxygen for providing heme to H. influenzae, strain Ela was grown in HDM containing myoglobin or hemoglobin as the sole source of heme. Growth of H. influenzae with myoglobin or hemoglobin as the sole source of heme was indistinguishable from that in the presence of free heme. In addition to heme proteins which contain heme as a functional ligand, hemopexin and haptoglobin are serum proteins which avidly bind and transport heme and hemoglobin, respectively. Heme bound to hemopexin (50% saturated) was readily available for growth of strain E1a in contrast to free heme. Similarly, the heme moiety of hemoglobin was readily available, even when bound to haptoglobin (50% saturated). Heme acquisition of strain Ela in plate assays. Heme acquisition by strain Ela in the plate assays was similar to that in broth (Fig. 4). Hemoglobin provided a readily available source of heme for growth of strain Eia. Binding heme and hemoglobin to hemopexin and haptoglobin did not prevent growth of strain Ela. Binding hemoglobin to haptoglobin decreased the growth zone diameter, although dense bacterial growth was evident. This difference in zone diameter may be due to reduced diffusion of the hemoglobinhaptoglobin macromolecular complex compared with hemoglobin alone. The relative lack of availability of heme from catalase and lactoperoxidase was confirmed in the HDMagar assay; growth zones were not detectable around wells containing 0.1,ug of heme as lactoperoxidase or catalase. However, a small zone of growth was detected around the well containing a 10-fold greater quantity of catalase; in contrast, no zone of growth was detected around the well containing a 10-fold greater quantity of lactoperoxidase. Heme acquisition by other strains of H. influenzae. To determine whether strain Ela was unique in its pattern of heme acquisition, untypable strains and type b strains of H. influenzae isolated from mucosal surfaces of healthy children were compared with invasive type b isolates (Table 1). All strains acquired heme from hemoglobin. Binding heme to hemopexin or albumin increased the growth zone diameter, possibly because of associated increased solubility. Inducible outer membrane proteins. As an initial step in determining possible mechanisms of acquisition of heme, we compared the outer membrane protein profile of strain Ela grown in limiting concentrations of heme (0.005 p.giml) or heme-replete conditions (10,ug/ml). A heme-repressible TABLE 1. Heme acquisition by different H. influenzae strainsa Colony diameter (mm) with: Strain Source Free (type) heme Heme-Hpx Heme-HSA Hgb Hgb-Hpt Ela (b) CSF C65 (b) CSF C148 (b) CSF R152 (b) CSF C120 (b) RT C132 (u) Eye C134 (u) RT C136 (u) NP R211 (b) Throat R212 (b) Throat R213 (b) Throat R214 (b) Throat a Each well contained 0.1,ug of heme in the form indicated. Abbreviations: Hpx, hemopexin; Hgb, hemoglobin; Hpt, haptoglobin; HSA, human serum albumin; RT, respiratory tract; NP, nasopharynx.

5 152 STULL A B 45 4-_ 2 1 t-n FIG. 5. SDS-PAGE of outer membrane proteins of strain Ela after growth in HDM containing (lane A) replete (10,ug/ml) or (lane B) growth-restricting (0.005 p.g/ml) concentrations of heme. The arrow indicates a heme-repressible outer membrane protein with an apparent molecular mass of 38 kda. The molecular mass standards were the same as those in Fig. 1. outer membrane protein with an apparent molecular mass of 38 kilodaltons (kda) was detectable in heme-restricted organisms compared to heme-replete organisms (Fig. 5). DISCUSSION The role of heme sources in the pathogenesis of invasive disease due to Hib has not previously been investigated. The presence of hemoglobin markedly enhances the virulence of E. coli, although hemoglobin bound to haptoglobin is unavailable (8). In these experiments, iron appeared to function as the critical moiety. The role of hemopexin, albumin, and other heme proteins in bacterial pathogenesis is poorly understood. Although H. influenzae requires heme for growth, the quantity of heme required depends on the aerobic metabolic activity of the organism. H. influenzae can be passaged several times in enriched media without added heme when incubated anaerobically (10). In contrast, growth is inhibited during the first aerobic passage without heme supplementation. Based on growth zone diameters, the heme requirements of invasive isolates could not be differentiated from noninvasive isolates. The source of heme for Hib during invasive infections is unknown. Of the heme proteins investigated in vitro, myoglobin and hemoglobin were the most accessible sources of heme for Hib. These data are limited by the use of animal sources for certain heme proteins, which may not be identical to their human counterparts. However, whale myoglobin did not restrict acquisition of heme by Hib. Although hemoglobin and myoglobin are primarily intracellular proteins, invasive infections due to Hib have been associated with extracellular hemoglobin due to hemolytic anemia (13). Shurin et al. (24) have proposed that the polyribosyl ribitol phosphate capsule of Hib may be absorbed to erythrocytes, leading to immune-mediated destruction of erythrocytes. Hemoglobin release may provide a source of heme for Hib during invasive infections. In response to restriction of certain nutrients, such as iron, other gram-negative organisms may express new outer membrane proteins which function as receptors for uptake mechanisms (2). Coulton and Pang (5) reported a heme-repressible outer membrane protein with an apparent molecular mass of 43 kda in one strain of H. influenzae, ATCC The finding of a heme-repressible outer membrane protein with a molecular mass of 38 kda in strain Ela may indicate that these proteins are not involved in heme-specific transport but are expressed during energy starvation. With in vivo labeled Fe-55 heme and hemoglobin, heme acquisition of strain Ela was not detectably altered by heme starvation (data not shown). Alternatively, outer membrane proteins with similar functions may have different apparent molecular masses within the species. Serum proteins which bind iron and iron-containing compounds may function in host defense by starving invasive pathogens for this essential element (3). E. coli produces low-molecular-weight iron-binding compounds called siderophores during iron starvation, and certain siderophores are effective in removing iron from transferrin (7, 11, 14). Hemoglobin is also an efficient source of iron for E. coli. However, binding hemoglobin with haptoglobin effectively sequesters the iron and, presumably, the heme from E. coli. The above data indicate that both invasive and noninvasive isolates of H. influenzae efficiently remove heme from hemoglobin bound to haptoglobin. Thus, haptoglobin may not function as a host defense factor against H. influenzae. The ability of hemopexin or albumin to sequester heme or iron from bacteria has not been previously investigated. The above data indicate that heme bound to hemopexin or albumin is readily available to H. influenzae. Herrington and Sparling (12) demonstrated that iron-transferrin can provide the iron necessary for growth of H. influenzae; it is not yet clear whether acquisition of heme from the above sources will meet the total iron requirements of H. influenzae. These data raise important questions concerning the molecular mechanisms of acquisition of heme which is avidly bound to proteins. Future studies will focus on acquisition of iron within the heme moiety and the genetics of iron acquisition in H. influenzae. ACKNOWLEDGMENTS INFECT. IMMUN. This work was supported in part by New Investigator Award R23 Al from the National Institutes of Health. The technical assistance of Kenneth Smith, James Wooten, and Barbara Daley is gratefully acknowledged. I also thank Michelle O'Donnell for excellent secretarial assistance. LITERATURE CITED 1. Arcoleo, J. P., and J. Greer Hemoglobin binding site and its relationship to the serine protease-like active site of haptoglobin. J. Biol. Chem. 257: Braun, V., R. E. W. Hancock, K. Hantke, and A. Hartmann Functional organization of the outer membrane of Escherichia coli: phage and colicin receptors as components of iron uptake systems. J. Supramol. Struct. 5: Bullen, J. J The significance of iron in infection. Rev. Infect. Dis. 3: Catlin, B. F Nutritional profiles of Neisseria gonorrhoeae, Neisseria meningitidis, and Neisseria lactimica in chemically defined media and the use of growth requirements for gonococcal typing. J. Infect. Dis. 128: Coulton, J. W., and J. C. S. Pang Transport of hemin by

6 VOL. 55, 1987 Haemophilus influenzae type b. Curr. Microbiol. 9: Delers, F., C. Lombart, M. Domingo, and S. Musquera A novel and specific method for the purification of hemoglobinbinding proteins. Anal. Biochem. 118: delorenzo, V., A. Bindereif, B. H. Paw, and J. B. Neilands Aerobactin biosynthesis and transport genes of plasmid ColV-K30 in Escherichia coli K-12. J. Bacteriol. 165: Eaton, J. W., P. Brandt, and J. R. Mahoney Haptoglobin: a natural bacteriostat. Science 215: Fildes, P The nature of the effect of blood-pigment upon the growth of B. influenzae. Br. J. Exp. Pathol. 2: Gilder, H., and S. Granick Studies on the Haemophilus group of organisms. J. Gen. Physiol. 31: Hartmann, A., and V. Braun Iron uptake and iron limited growth of Escherichia coli K-12. Arch. Microbiol. 130: Herrington, D. A., and P. F. Sparling Haemophilus influenzae can use human transferrin as a sole source for required iron. Infect. Immun. 48: Kaplan, K. M., and F. A. Oski Anemia with Haemophilus influenzae meningitis. Pediatrics 65: Konopka, K., A. Bindereif, and J. B. Neilands Aerobactin-mediated utilization of transferrin iron. Biochemistry 21: Koskelo, P., and U. Muller-Eberhard Interaction of porphyrins with proteins. Semin. Hematol. 14: LaScolea, L. J., S. V. Rosales, R. C. Welliver, and P. L. Ogra Mechanisms underlying the development of meningitis or epiglottitis in children after Haemophilus influenzae type b bacteremia. J. Infect. Dis. 151: Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193: Mueller-Eberhard, U., and W. T. Morgan Porphyrinbinding proteins in serum. Ann. N.Y. Acad. Sci. 244: Palumbo, G., and M. F. Tecce A four- to six-fold enhancement in sensitivity for detecting trace proteins in dye or silver stained polyacrylamide gels. Anal. Biochem. 134: Riggs, A Preparation of blood hemoglobin of vertebrates. Methods Enzymol. 76: Rubin, L. G., A. Zwahleu, and E. R. Moxon Role of intravascular replication in the pathogenesis of experimental bacteremia due to Haemophilus influenzae type b. J. Infect. Dis. 152: Seery, V. L., G. Hathaway, and U. Muller-Eberhard Hemopexin of human and rabbit: molecular weight and extinc- HEME SOURCES FOR H. INFLUENZAE 153 tion coefficient. Arch. Biochem. Biophys. 150: Seery, V. L., and U. Muller-Eberhard Binding of porphyrins to rabbit hemopexin and albumin. J. Biol. Chem. 248: Shurin, S. B., P. Anderson, J. Zollinger, and R. K. Rathbun Pathophysiology of hemolysis in infections with Haemophilus influenzae type b. J. Clin. Invest. 77: Smith, A. L., R. S. Daum, D. Scheifele, V. Syriopolou, D. R. Averill, M. D. Roberts, and T. L. Stull Pathogenesis of Haemophilus influenzae meningitis, p In S. H. Sell and P. F. Wright (ed.), Haemophilus influenzae: epidemiology, immunology and prevention of disease. Elsevier Biomedical Press, Amsterdam. 26. Smith, A. L., D. H. Smith, D. R. Averill, Jr., J. Marino, and E. R. Moxon Production of Haemophilus influenzae b meningitis in infant rats by intraperitoneal inoculation. Infect. Immun. 8: Stull, T. L., R. F. Jacobs, J. E. Haas, M. C. Roberts, C. B. Wilson, and A. L. Smith Human serum bactericidal activity against Haemophilus influenzae type b. J. Gen. Microbiol. 130: Stull, T. L., K. Mack, J. E. Haas, J. Smit, and A. L. Smith A comparison of techniques for isolation of the outer membrane proteins of Haemophilus influenzae type b. Anal. Biochem. 150: Stull, T. L., P. M. Mendelman, J. E. Haas, M. A. Shoenborn, K. D. Mack, and A. L. Smith Characterization of Haemophilus influenzae type b fimbriae. Infect. Immun. 46: Thomas, P. E., D. Ryan, and W. Levin An improved staining procedure for the detection of the peroxidase activity of cytochrome P-450 on sodium dodecyl sulfate polyacrylamide gels. Anal. Biochem. 75: Tsutusi, K., and G. C. Mueller Affinity chromatography of heme-binding proteins: an improved method for the synthesis of hemin-agarose. Anal. Biochem. 121: Turk, D. C Clinical importance of Haemophilus influenzae-1981, p In S. H. Sell and P. F. Wright (ed.), Haemophilus influenzae: epidemiology, immunology, and prevention of disease. Elsevier Biomedical Press, Amsterdam. 33. Wejman, J. C., D. Hovsepian, J. S. Wall, J. F. Hainfeld, and J. Greer Structure of haptoglobin and the haptoglobinhemoglobin complex by electron microscopy. J. Mol. Biol. 174: Wray, W., T. Boulikas, V. P. Wray, and R. Hancock Protein silver staining in polyacrylamide gels. Anal. Biochem. 118: