2007-Role of P-Fimbrial-mediated Adherence In

8/12/2019 2007-Role of P-Fimbrial-mediated Adherence In

1/7

Role of P-fimbrial-mediated adherence inpyelonephritis and persistence of uropathogenic

Escherichia coli(UPEC) in the mammalian kidneyMC Lane1 and HLT Mobley1

1Department of Microbiology and Immunology, University of Michigan Medical School, Ann Arbor, Michigan, USA

P fimbria, a mannose-resistant adhesin of uropathogenic

Escherichia coli(UPEC), has been shown to be associated with

acute pyelonephritis. Thepapgene cluster encodes the

proteins required for P-fimbrial biogenesis, including papG,

which encodes the tip adhesin. The three most studied PapG

molecular variants, which are shown to bind distinctisoreceptors, are PapGI, -II, and -III. PapGII preferentially binds

globoside, or GbO4, a glycolipid isoreceptor of the human

kidney. Studies using different animal models of ascending

urinary tract infection (UTI) have demonstrated a variable

role for P fimbriae, and specifically PapGII-mediated

adherence, in renal colonization. The disparities in the results

obtained from those studies are likely to be attributed to the

differences in animal models and UPEC strains utilized. One

explanation that is discussed in detail is the contribution of

multiple fimbriae of UPEC that potentially mediate adherence

to the mammalian kidney. Overall, P fimbriae appear to play

some role in mediating adherence to uroepithelial cells invivoand establishing an inflammatory response during renal

colonization, thus contributing to kidney damage during

acute pyelonephritis. To verify that P fimbriae contribute to

the pathogenesis of UPEC during ascending UTI (and in

particular acute pyelonephritis), future studies should be

conducted to satisfy fully all three tenets of the molecular

Kochs postulates, including complementation of a mutated

allele.

Kidney International (2007) 72, 1925; doi:10.1038/sj.ki.5002230;

published online 28 March 2007

KEYWORDS: uropathogenic E. coli; UPEC; P fimbriae; PapG; adherence;

pyelonephritis

While the urinary tract of a healthy adult is generally sterile,it is also the most common site of bacterial infection. Ingeneral, urinary tract infections (UTIs) develop in anascending manner beginning with peri-urethral colonization,followed by ascension of the urethra into the bladder causing

cystitis, and in some cases if left untreated, ascension of theureters into the kidneys establishing acute pyelonephritis.Acute pyelonephritis is considered the most serious UTI as itis known to typically cause scarring of the kidney, which mayin turn lead to irreversible kidney damage, kidney failure,and/or sepsis. In more than 80% of all uncomplicated acutepyelonephritis cases, the etiologic agent is uropathogenicEscherichia coli(UPEC).1

An important stage in the successful colonization of theurinary tract and pathogenesis of UTI is the ability of UPECto adhere to host uroepithelia. In 1976, Eden et al .2

demonstrated that E. coli isolated from the urine of patients

with acute symptomatic pyelonephritis adhered in greaternumbers to exfoliated uroepithelial cells than E. coli isolatedfrom the urine of patients with asymptomatic bacteriuria.Two years later, the ability of UPEC to attach to humanuroepithelial cells was attributed to the presence of fimbriae(or pili), which appear as hair-like appendages that protrudefrom the surface of bacteria.3 These fimbriae were determinedto be distinct from the common fimbriae (otherwise knownas type 1 fimbriae) in that they mediated adherence touroepithelial cells in the presence of mannose, a knowninhibitor of type 1-mediated adherence.3

The most extensively studied adhesin, and also the firstvirulence-associated factor identified for UPEC is P fimbria.

P fimbriae, encoded by the pap (pyelonephritis-associatedpili) genes, are significantly prevalent among strains of UPECthat cause pyelonephritis4 and are characterized by theirmannose-resistant adherence to Gal(a14)Galb moietiespresent in the globoseries of membrane glycolipids onhuman erythrocytes of the P blood group and onuroepithelial cells5,6 (for a good review on the discoveryand characterization of P fimbriae, see Johnson7 and Joneset al.8). Structurally speaking, they are composed ofB1000copies of the major subunit protein, PapA, which polymerizeto form a rigid stalk that is connected to a flexible tipconsisting of limited copies of the minor subunit proteins,

http://www.kidney-international.org m i n i r e v i e w

& 2007 International Society of Nephrology

Received 2 February 2007; accepted 13 February 2007; published online

28 March 2007

Correspondence: HLT Mobley, Department of Microbiology and Immuno-

logy, University of Michigan Medical School, 5641 Medical Science Building II,

1150 West Medical Center Drive, Ann Arbor, Michigan 48109, USA.

E-mail: [email protected]

Kidney International(2007) 72 , 1925 19


2/7

PapE and PapF, and receptor-binding adhesin, PapG, at thedistal end.9,10

THE THREE CLASSES OF PapG ADHESINS AND THEIR

PREVALENCE AMONG DIFFERENT UPEC ISOLATES

P fimbriae are chromosomally encoded by the pap genecluster, consisting of 11 genes, including papG, whichencodes the P-fimbrial adhesin. Upon characterization ofthe PapG adhesin of UPEC isolate J96, Lund et al.11 identifiedan additional structurally related fimbrial gene cluster, prs

(pap-related sequence). While Prs fimbriae were shown to beserologically identical to P fimbriae (of the F13 serotype),they displayed a different receptor-binding specificity.11 Thedifferences in receptor-binding specificity were attributed tothe adhesins, PapG and PrsG; in particular, trans-comple-mentation of a P-fimbrial adhesin mutant (DpapG) withprsGaltered the binding specificity to that of Prs fimbriae.11

Two years later, Stromberg et al.12 conducted a studycomparing the binding specificities of PapG and PrsG of

strain J96, and PapG from UPEC isolates IA2 and AD110 (ofthe F72 and F11 P-fimbrial serotypes, respectively). Specifi-cally, the four G adhesins were examined for bindingto purified glycolipids and mammalian erythrocytes anduroepithelial cells. From those studies, they determined thatthe four G adhesins represent three different isoreceptor-binding variants (PapGI, -II, and -III). The isoreceptorsbound by the different PapG variants all contain a commonGal(a14)Gal moiety linked to a ceramide group, whichanchors the receptor in the lipid bilayer.13 However, theisoreceptors vary in the number of N-acetylgalactosaminemoieties added to the distal Gal(a14)Gal core, or by the

addition of sialic acid residues to form more complexreceptors.14,15 The type and distribution of the particularisoreceptors among different host tissues and the differentbinding preferences of the three PapG variants have beenshown to contribute to the differences in host tropism ofP-fimbriated E. coli.12

It was originally described that PapGI adhesins preferen-tially bind globotriaosylceramide or GbO3 (abundant onhuman uroepithelial cells), PapGII adhesins preferentiallybind globoside or GbO4 (abundant on human uroepithelialcells), and PapGIII adhesins (or PrsG) preferentiallybind Forssman antigen or GbO5 (predominant on canine,but not human uroepithelial cells).14 However, additional

studies have demonstrated that the PapG variants, inparticular PapGIII, bind to other isoreceptors present inthe urinary tract of humans.1517 Recently, additional minorPapG variants (of PapGI) as well as a new class of PapGvariants (PapGIV) have been reported.18,19 The isoreceptorsfor which these new PapG variants bind is currentlyunknown. Therefore, as we discover more isoreceptors andmore major or minor PapG variants, we can better under-stand the host tropism and how it relates to a particular PapGvariant.

Clinically, the class II papG allele is primarily associatedwith human pyelonephritis and bacteremia and the class III

papG allele is associated with human cystitis and withgenitourinary infections in dogs and cats.2025 Interestingly,the prevalence of the class II papG allele decreases in UPECstrains isolated from adults with acute pyelonephritis andurinary tract abnormalities and/or obstruction.26 Owing tothe rarity of the class I pap allele, not much is known aboutits clinical association.

The class of PapG variant and the chromosomal locationof pap alleles typically differ among UPEC strains. The papgene clusters reside within pathogenicity islands that areacquired by horizontal transfer and inserted within oradjacent to specific tRNA genes. For example, in thepyelonephritis isolates, CFT0732729 and J96,30,31 two papgene clusters reside in separate pathogenicity islands near thepheUand pheV tRNA genes. Interestingly, both papG genesfrom CFT073 encode the PapGII adhesin variant, while thetwo papG genes from J96 encode the PapGI and PapGIIIadhesin variants. UPEC cystitis isolate UTI8932,33 and UPECpyelonephritis isolate 53634,35 have a single pap gene cluster,

which is contained in one pathogenicity island that isinserted near the leuX tRNA gene. Both papG genes fromUTI89 and 536 encode the PapGIII adhesin variant. UPECbacteremia isolate, CP9 (J96-like) has been shown to containtwo papG alleles that encode the PapGI and PapGIIIvariants.36 Lastly, UPEC bacteriuria isolate IA2 and pyelone-phritis isolates AD110 and DS17 all contain a single pap genecluster that encodes the PapGII variant.12,3739

Figure 1 illustrates the conservation of amino-acidsequences between the different PapG variants. Each of thePapG adhesin sequences from UPEC strains J96, CP9, IA2,CFT073, DS17, 536, and UTI89 were obtained from Entrez

PubMed and aligned using ClustalW software (MegAlign,Lasergene 7). It is hypothesized that the differences in thePapG amino-acid sequences allow for alterations in adhesinbinding and host specificity. Recently, both the solutionstructure of the PapGII adhesin domain and the crystalstructure of the PapGII receptor bound to GbO4 (predomi-nant isoreceptor of the human kidney) as well as theunbound form of the adhesin have been determined usingmultidimensional nuclear magnetic resonance and themultiwavelength anomalous dispersion phasing method.40,41

In Figure 1, residues of PapGII that have been shown to beinvolved in GbO4 receptor binding are boxed in red.40 Usinga series of sophisticated modeling programs, Dodson et al.40

were also able to predict the residues of PapGIII that makecontact with GbO5; in Figure 1, for comparison with thePapGII-binding residues, the predicted contact residues ofPapGIII are boxed in green. This highlights the structuralconservation among PapG variants, while showing thedivergence leading to altered receptor specificities.

CONTRIBUTION OF P-FIMBRIAL-MEDIATED ADHERENCE TO

PYELONEPHRITIS AND THE PERSISTENCE OF UPEC IN THE

MAMMALIAN KIDNEY

Since the discovery of P fimbriae, it has been hypothesizedthat these adhesins contribute to the pathogenesis of UPEC

20 Kidney International(2007) 72 , 1925

m i n i r e v i e w MC Lane and HLT Mobley: P-fimbrial-mediated adherence of UPEC


3/7

within the mammalian urinary tract. An earlier study,conducted in 1987, demonstrated that the serum of femalepatients with symptoms of pyelonephritis contained P-fimbrial antibodies, suggesting that P fimbriae were expressedduring infection.42 Similarly, another study conductedshortly thereafter demonstrated that bacteria obtained frommidstream or catheterized urine specimens from patientswith E. coli cystitis expressed type 1 and P fimbriae.43 Thus,

both studies provided compelling evidence for the in vivoexpression of P fimbriae during human UTI.

To determine precisely the contribution of P fimbriae tothe pathogenesis of UPEC, additional studies have beenconducted using different animal models of infection. In1985, OHanley et al.44 determined the distribution anddensity of the P-fimbrial receptor within the urinary tract ofBALB/c mice to elucidate the molecular basis for renalcolonization. Using the mouse model of ascending infection,they demonstrated that a Gal-Gal pili producing transfor-mant of the nonpathogenicE. colistrain HB101 (obtained bycloning random segments of a UPEC strain J96 cosmid

library into a moderate copy plasmid) was able to colonizethe kidneys of BALB/c mice in the absence of vesicoureteralreflux unlike the nonpiliated HB101 parent strain.44 More-over, subcutaneous and intramuscular vaccination of themice with purified Gal-Gal pili before infection could blockrenal colonization upon infection with the Gal-Gal piliproducing UPEC strain J96.44 A similar study, conducted byHagberg et al.45 in 1983, also utilized UPEC strain J96 as a

donor to transform fecal-commensal E. coli strain 506 withthe pap gene cluster. In particular, they demonstrated thattransformants of 506 that contained the pap gene clusterpersisted within the kidneys of CBA and BALB/c mice moreefficiently than did the pap-negative transformants.45 It isimportant to note that in both studies, wild-type UPEC levelsof pyelonephritis were never achieved with the commensalE. coli pap transformants, suggesting that factors other thanP fimbriae are needed for optimal renal colonization.Furthermore, while both studies demonstrate that additionof P fimbriae to a commensal E. coli strain can allow forbetter colonization, they do not prove that P fimbriae are

Figure 1 | Amino-acid sequence alignment of the three major PapG (or PrsG) adhesin variants from UPEC strains J96, CP9, CFT073, IA2,DS17, 536, and UTI89. The amino-acid sequences, obtained from Entrez PubMed, are presented in single-letter code. The different PapGvariants are divided by a solid line. Residues that are homologous to the consensus (which is not shown) are shaded in yellow. Residues that arenot homologous to the consensus are shaded in blue. Residues of PapGII that have been shown to be involved in GbO4 receptor binding areboxed in red.40 Residues of PapGIII that have been predicted to be in contact with GbO5 are boxed in green.40


MC Lane and HLT Mobley: P-fimbrial-mediated adherence of UPEC m i n i r e v i e w


4/7

required for UPEC pathogenesis or that they contribute tothe fitness of UPEC during UTI.

To determine whether P fimbriae are indispensable forUPEC pathogenesis, isogenic P-fimbrial mutants of differentUPEC strains have been constructed and studied in differentanimal models of ascending UTI. The first study toinvestigate the fitness of an isogenic P-fimbrial mutantduring ascending UTI was conducted by Hagberg et al.,45

along with the P-fimbrial transformant studies as describedpreviously. In particular, they demonstrated that a UPECstrain lacking P fimbriae was significantly outcompeted bywild-type UPEC in the kidneys of CBA and BALB/c miceduring mixed infection (or co-challenge).45 However, mu-tants in that study were generated with the use of chemicalmutagenesis; since chemical mutagenesis often causessecondary mutations, it would be impossible to knowwhether the outcompetition of the P-fimbrial UPEC mutantwas solely due to loss of P fimbriae or some other UPECfactor that was mutagenized. Several years later, our

laboratory utilized allelic exchange mutagenesis to createprecise deletions of papEFG (pheV copy) and papDEFG(pheU copy), thus creating a double pap mutant of UPECstrain CFT073 (UPEC76) that does not produce P fimbriae.46

Upon transurethral inoculation of CBA mice (N 100) withfour different inoculum concentrations (105, 106, 107, and109 CFU) of either CFT073 or UPEC76, it was demonstratedthat after 1 week of infection, no significant differences inorganism concentration or histological findings between thewild-type and double-papmutant were detected in the urine,bladder, or kidney at any challenge concentration. Thus, itwas concluded that P fimbriae of UPEC are not required for

ascending UTI of CBA/J mice. A primate model of ascendinginfection has also been used to study the contribution of Pfimbriae to UPEC pathogenesis, as monkeys are also knownto have the PapG adhesin receptors.47 In 1994, Robertset al.39

demonstrated that an isogenic papG mutant of UPEC strainDS17 (called DS17-8), created by allelic exchange mutagen-esis, was equally able to cause bladder infection ofcynomolgous monkeys as compared to wild-type DS17, butwas unable to cause acute pyelonephritis such as wild-typeDS17 in this model as determined by histopathologicalcriteria and the concentration of the DS17-8 mutant withinthe kidneys. Although the PapG adhesin mutant was shownnot to cause pyelonephritis, the establishment of P fimbriae

as a virulence factor, according to molecular Kochspostulates, has still not been fully satisfied because of thelack of complementation of the mutant in vivo. In all, theinconsistencies observed between each of the P-fimbrialmutant animal studies could be explained by the differencesin animal models or UPEC strains utilized. Although bothmice and monkeys are known to have receptors for the PapGadhesin, a study conducted in 1995 by Lanne et al.48 hasshown that the isoreceptor patterns present within thekidneys of primates and mice differ significantly, thuspossibly contributing to differences in host specificity. Also,different PapG variants expressed by the UPEC strains

studied (PapGI and -III of J96 and PapGII of CFT073 andDS17) may also result in altered host specificity. Anotherexplanation could include differences in the host immuneresponse to P fimbriae, thus allowing some animal strains tobe less susceptible than others to P-fimbriated UPECinfection. Lastly, because it is known that UPEC may havethe capacity to express up to 12 different fimbriae, it isconceivable that loss of P-fimbrial-mediated adherence andcolonization may be compensated for by the expression ofanother fimbrial type (discussed further in the next section).

More recent studies have uncovered a molecular crosstalkbetween innate immune Toll-like receptor 4 binds bacteriallipopolysaccharide signaling and P-fimbrial-mediated attach-ment, which is lipopolysaccharide-independent (for a goodreview see Bergsten49). Upon P-fimbrial attachment to itsglycosphingolipid receptor, ceramide is released from thelipid part of the receptor; in particular, it has recently beenshown that ceramide acts as an agonist of Toll-like receptor 4and potentially acts as a signaling intermediate between Toll-

like receptor 4 and the glycosphingolipid receptor.50 Activa-tion of the Toll-like receptor 4 receptor by P-fimbrialattachment subsequently leads to the production of proin-flammatory cytokines and chemokines (interleukin-6 andCXCL8, respectively) and recruitment of neutrophils.49

Although this proinflammatory response is beneficial ininitiating bacterial clearance, it also causes damage to thesurrounding tissue and is associated with renal complica-tions. Since P fimbriae are implicated in triggering inflam-mation, it can be deduced that they may also contribute tothe pathology and symptoms of acute pyelonephritis.

REDUNDANCY OF FIMBRIAL SYSTEMS IN UPECUpon sequencing of the different UPEC genomes, a numberof distinct fimbrial gene clusters have been identified. Forexample, the sequencing of UPEC strain CFT073 has revealed12 putative fimbrial gene clusters.29 The most extensivelystudied fimbriae that have been shown to be important inurinary tract colonization are the type 1, P, S, F1C, andDr/AFA adhesins. Each fimbrial type binds distinct isorecep-tors found within the mammalian urinary tract: type 1fimbriae bind N-linked oligomannose glycoproteins anduroplakin,51 P fimbriae bind Gala(14)bGal moieties presentin the globoseries of glycolipids,5,6 S fimbriae bind a-sialyl-2,3-b-galactoside-containing receptors,52,53 F1C fimbriae

bind lactosylceramide-containing receptors,54,55 and Dr/Afaadhesins bind the Dr (a ) blood group antigen present onthe complement cascade regulator factor, decay-acceleratingfactor.56 Although these fimbriae recognize different recep-tors, each of the receptors may occupy sites within the sametissue. For instance, type 1, P, S, F1C, and Dr fimbriae are allknown to bind to different sites within the human kidney(Figure 2). Figure 2 illustrates the sites within the humankidney where distinct fimbrial types have been shownexperimentally to bind.57 The fact that multiple fimbrialtypes are able to bind to various sites of the human kidneysuggests that UPEC are provided with multiple means to




5/7

establish renal adherence, thus, contributing to the patho-gens overall success during renal colonization.

For the most part, E. coli express one fimbrial type at atime.58 Thus, it is not surprising that crosstalk occursbetween the regulators of the different fimbrial systems ofUPEC (for a good review see Holden59). Recently, ourlaboratory and others have demonstrated that loss of onefimbrial type leads to expression of another in UPEC.

Specifically, our laboratory showed that a mutant of UPECstrain CFT073 unable to make type 1 and P fimbriaeproduces F1C fimbriae, providing a compensatory mechan-ism by which UPEC are still able to maintain adherence.60

This compensatory mechanism has also been observed forUPEC strain UTI89, in which a mutant unable to make type1 pili upregulated expression of S pili (Wright KJ, Seed PC,Hultgren SJ, Abstr. 105th Gen. Meet. Am Soc Microbiol2005;abstract. B-176: 116). To provide an example, Figure 3demonstrates that a type 1 fimbrial phase-locked derivative ofCFT073 that is unable to produce type 1 fimbriae (fimL-OFF), expresses other fimbriae (panels C and D). The type1 fimbrial phase-locked derivative of CFT073 that constitu-

tively expresses type 1 fimbriae (fim L-ON), which wascultured statically as was the fim L-OFF derivative, is shownfor comparison (panels A and B). Together, these datademonstrate that there exists a redundancy of fimbrialsystems in UPEC, which provides UPEC with multiplealternatives to maintain adherence during UTI. Thus, aspreviously discussed, the lack of importance for P fimbriae inthe CBA mouse model of infection46 could be explained bythe redundancy of fimbrial systems in UPEC strain CFT073(which produces several fimbriae, including type 1, P, F1C,and Dr fimbriae). In view of the data summarized in Figure 2,it is possible that type 1, F1C, or Dr fimbriae may be

contributing to the colonization of the murine renalepithelium in the absence of P-fimbrial expression.

SUMMARY, CONCLUSIONS, AND FUTURE PERSPECTIVES

P fimbriae, the best-studied fimbriae and first virulence-associated factor described for UPEC, have been shown to be

Figure 2 | Cross-section of the human kidney displaying UPEC fimbrial adhesin-binding sites. Pyelonephritis is considered an ascendinginfection, where UPEC from the bladder ascend through the ureters into the kidneys. Following ascension, UPEC are known to colonize thecollecting ducts, distal and proximal tubules, glomeruli, Bowmans capsules, and the blood vessel walls (see inset). Specific UPEC fimbrialadhesins that have been shown to be involved in attachment to these sites within the kidney are labeled in blue.57

a

c d

b

Figure 3 | Transmission electron micrographs of UPEC expressing

different fimbriae. (aand b) CFT073 fim L-ON, a mutant thatconstitutively expresses type 1 fimbriae. (candd) CFT073fimL-OFF, amutant that is unable to express type 1 fimbria produces anothertype of fimbriae. a and c are at 34 000 magnification, andb and dare at 64 000 magnification.




6/7

associated with acute pyelonephritis. The pap gene clusterencodes the proteins required for P-fimbrial synthesis,including papG, the gene encoding the tip fimbrial adhesin.Three major and well-studied classes of papG alleles exist,which encode the molecular variants PapGI, -II, and -III.Each PapG variant is known to have a distinct isoreceptorspecificity, which in turn results in altered host tissuetropism. PapGII, which is clinically associated with acutepyelonephritis in humans, has been shown to bindpreferentially globoside, or GbO4, the predominant glycoli-pid isoreceptor of the human kidney. Various research usingdifferent animal models of ascending UTI have found avariable role for P fimbriae, and in particular PapGII-mediated adherence, in the colonization of the mammaliankidney. The disparities in the results obtained from thosestudies are likely to be attributed to the differences in animalmodels and UPEC strains utilized. One likely explanationthat was discussed in detail was the fact that UPEC produceseveral different fimbrial types that are also known to bind to

different regions of the human kidney, and therefore, when Pfimbriae are unable to be expressed, other fimbriae couldcompensate for their loss. Overall, it appears that there is asubtle role for P fimbriae in mediating adherence touroepithelial cells in vivo and establishing a robust inflam-matory response during renal colonization, which in turncontributes to kidney damage during acute pyelonephritis. Toverify that P fimbriae contribute to the pathogenesis of UPECduring ascending UTI (and in particular acute pyelone-phritis), future studies should be conducted to satisfy fully allthree tenets of the molecular Kochs postulates, includingcomplementation of a defined mutation. Furthermore, it may

be advantageous to investigate the spatial and temporalexpression of P fimbriae (and other fimbrial types) during invivoascending infection to determine more precisely the roleand expression profile of P fimbriae (and other fimbriae)during kidney colonization.

ACKNOWLEDGMENTS

We sincerely apologize to the authors of publications that could notbe included in this review due to space limitations. This work wassupported by Public Health Service grant AI43363 and AI059722 fromthe National Institutes of Health.

REFERENCES1. Stamm WE, Hooton TM. Management of urinary tract infections in adults.

N Engl J Med1993; 329: 13281334.2. Eden CS, Hanson LA, Jodal U et al.Variable adherence to normal human

urinary-tract epithelial cells of Escherichia colistrains associated withvarious forms of urinary-tract infection.Lancet1976; 1 : 490492.

3. Eden CS, Hansson HA. Escherichia colipili as possible mediators ofattachment to human urinary tract epithelial cells.Infect Immun1978;21:229237.

4. Kallenius G, Mollby R, Svenson SB et al. Occurrence of P-fimbriatedEscherichia coliin urinary tract infections.Lancet1981; 2: 13691372.

5. Leffler H, Svanborg-Eden C. Glycolipid receptors for uropathogenicEscherichia colion human erythrocytes and uroepithelial cells. InfectImmun 1981;34: 920929.

6. Leffler H, Svanborg-Eden C. Chemical identification of aglycosphingolipid receptor for Escherichia coliattaching to humanurinary tract epithelial cells and agglutinating human erythrocytes. FEMSMicrobiol Lett1980; 8: 127134.

7. Johnson JR. Virulence factors inEscherichia coliurinary tract infection.ClinMicrobiol Rev1991; 4: 80128.

8. Jones CH, Dodson K, Hultgren SJ. Structure, function and assembly ofadhesive P pili. In: Mobley HLT, Warren JW (eds) Urinary Tract Infections:Molecular Pathogenesis and Clinical Management. ASM Press: Washington,D.C., 1996, pp 175219.

9. Kuehn MJ, Heuser J, Normark Set al. P pili in uropathogenic E. coliarecomposite fibres with distinct fibrillar adhesive tips.Nature 1992; 356:252255.

10. Lindberg F, Lund B, Johansson Let al. Localization of the receptor-binding protein adhesin at the tip of the bacterial pilus. Nature1987;328:8487.

11. Lund B, Marklund BI, Stromberg N et al. Uropathogenic Escherichia colican express serologically identical pili of different receptor bindingspecificities.Mol Microbiol1988; 2: 255263.

12. Stromberg N, Marklund BI, Lund B et al.Host-specificity of uropathogenicEscherichia colidepends on differences in binding specificity to Gal alpha14Gal-containing isoreceptors.Embo J1990; 9: 20012010.

13. Hakomori S. Bifunctional role of glycosphingolipids. Modulators fortransmembrane signaling and mediators for cellular interactions.J BiolChem1990; 265: 1871318716.

14. Stromberg N, Nyholm PG, Pascher I et al. Saccharide orientation at thecell surface affects glycolipid receptor function.Proc Natl Acad Sci USA1991; 88: 93409344.

15. Stapleton AE, Stroud MR, Hakomori SIet al. The globoseriesglycosphingolipid sialosyl galactosyl globoside is found in urinary tract

tissues and is a preferred binding receptor in vitro for uropathogenicEscherichia coliexpressingpap-encoded adhesins.Infect Immun1998;66:38563861.

16. Karr JF, Nowicki BJ, Truong LDet al. pap-2-encoded fimbriae adhere tothe P blood group-related glycosphingolipid stage-specific embryonicantigen 4 in the human kidney. Infect Immun 1990; 58: 40554062.

17. Lindstedt R, Baker N, Falk P et al. Binding specificities of wild-type andclonedEscherichia colistrains that recognize globo-A.Infect Immun1989;57: 33893394.

18. Johnson JR, Stell AL, Kaster Net al. Novel molecular variants of allele I ofthe Escherichia coliP fimbrial adhesin genepapG.Infect Immun 2001;69:23182327.

19. Manning SD, Zhang L, Foxman B et al. Prevalence of known P-fimbrial Galleles in Escherichia coliand identification of a new adhesin class. ClinDiagn Lab Immunol2001; 8 : 637640.

20. Johnson JR, OBryan TT, Low DAet al.Evidence of commonality betweencanine and human extraintestinal pathogenic Escherichia colistrains that

express papGallele III. Infect Immun 2000; 68: 33273336.21. Johnson JR, Russo TA, Brown JJet al. papG alleles of Escherichia colistrains causing first-episode or recurrent acute cystitis in adult women.

J Infect Dis 1998; 177: 97101.22. Otto G, Sandberg T, Marklund BIet al. Virulence factors and pap

genotype in Escherichia coliisolates from women with acutepyelonephritis, with or without bacteremia. Clin Infect Dis 1993; 17:448456.

23. Johanson IM, Plos K, Marklund BIet al. pap, papG and prsG DNAsequences in Escherichia colifrom the fecal flora and the urinary tract.Microb Pathog 1993; 15: 121129.

24. Johnson JR.papGalleles amongEscherichia colistrains causing urosepsis:associations with other bacterial characteristics and host compromise.Infect Immun 1998; 66: 45684571.

25. Johnson JR, Brown JJ, Maslow JN. Clonal distribution of the three allelesof the Gal(a14)Gal-specific adhesin genepapGamong Escherichia colistrains from patients with bacteremia.J Infect Dis 1998; 177: 651661.

26. Tseng CC, Huang JJ, Ko WCet al.Decreased predominance ofpapGclassII allele in Escherichia colistrains isolated from adults with acutepyelonephritis and urinary tract abnormalities. J Urol2001; 166:16431646.

27. Guyer DM, Kao JS, Mobley HL. Genomic analysis of a pathogenicity islandin uropathogenic Escherichia coli CFT073: distribution of homologoussequences among isolates from patients with pyelonephritis, cystitis, andcatheter-associated bacteriuria and from fecal samples. Infect Immun1998; 66: 44114417.

28. Rasko DA, Phillips JA, Li Xet al. Identification of DNA sequences from asecond pathogenicity island of uropathogenic Escherichia coliCFT073:probes specific for uropathogenic populations.J Infect Dis 2001; 184:10411049.

29. Welch RA, Burland V, Plunkett III Get al. Extensive mosaic structurerevealed by the complete genome sequence of uropathogenicEscherichia coli. Proc Natl Acad Sci USA 2002; 99: 1702017024.




7/7

30. Hull RA, Gill RE, Hsu Pet al.Construction and expression of recombinantplasmids encoding type 1 or D-mannose-resistant pili from a urinary tractinfection Escherichia coli isolate. Infect Immun 1981; 33: 933938.

31. Swenson DL, Bukanov NO, Berg DEet al. Two pathogenicity islands inuropathogenic Escherichia coliJ96: cosmid cloning and samplesequencing. Infect Immun 1996; 64: 37363743.

32. Mulvey MA, Schilling JD, Hultgren SJ. Establishment of a persistentEscherichia colireservoir during the acute phase of a bladder infection.Infect Immun 2001; 69: 45724579.

33. Chen SL, Hung CS, Xu Jet al. Identification of genes subject to positiveselection in uropathogenic strains ofEscherichia coli: a comparativegenomics approach.Proc Natl Acad Sci USA 2006; 103: 59775982.

34. Hacker J, Knapp S, Goebel W. Spontaneous deletions and flanking regionsof the chromosomally inherited hemolysin determinant of an EscherichiacoliO6 strain. J Bacteriol1983; 154: 11451152.

35. Middendorf B, Blum-Oehler G, Dobrindt Uet al.The pathogenicity islands(PAIs) of the uropathogenic Escherichia coli strain 536: island probing ofPAI II536. J Infect Dis 2001; 183(Suppl 1): S17S20.

36. Johnson JR, Russo TA, Scheutz Fet al.Discovery of disseminated J96-likestrains of uropathogenicEscherichia coliO4:H5 containing genes for bothPapG(J96) (class I) and PrsG(J96) (class III) Gal(alpha14)Gal-bindingadhesins. J Infect Dis 1997; 175: 983988.

37. Clegg S. Cloning of genes determining the production of mannose-resistant fimbriae in a uropathogenic strain ofEscherichia colibelongingto serogroup O6. Infect Immun 1982; 38: 739744.

38. van Die I, Spierings G, van Megen I et al. Cloning and genetic

organization of the gene cluster encoding F71 fimbriae of auropathogenic Escherichia coliand comparison with the F72gene cluster.FEMS Microbiol Lett1985; 28: 329334.

39. Roberts JA, Marklund BI, Ilver Det al.The Gal(a14)Gal-specific tip adhesinofEscherichia coliP-fimbriae is needed for pyelonephritis to occur in thenormal urinary tract.Proc Natl Acad Sci USA 1994; 91: 1188911893.

40. Dodson KW, Pinkner JS, Rose Tet al.Structural basis of the interaction ofthe pyelonephriticE. coliadhesin to its human kidney receptor. Cell2001;105: 733743.

41. Sung MA, Fleming K, Chen HAet al.The solution structure of PapGII fromuropathogenic Escherichia coliand its recognition of glycolipid receptors.EMBO Rep 2001; 2: 621627.

42. de Ree JM, van den Bosch JF. Serological response to the P fimbriae ofuropathogenic Escherichia coli in pyelonephritis. Infect Immun 1987; 55:22042207.

43. Kisielius PV, Schwan WR, Amundsen SKet al. In vivo expression andvariation ofEscherichia colitype 1 and P pili in the urine of adults with

acute urinary tract infections. Infect Immun 1989; 57: 16561662.44. OHanley P, Lark D, Falkow Set al. Molecular basis of Escherichia colicolonization of the upper urinary tract in BALB/c mice. Gal-Gal piliimmunization prevents Escherichia colipyelonephritis in the BALB/cmouse model of human pyelonephritis.J Clin Invest1985; 75: 347360.

45. Hagberg L, Hull R, Hull Set al. Contribution of adhesion to bacterialpersistence in the mouse urinary tract. Infect Immun 1983; 40:265272.

46. Mobley HL, Jarvis KG, Elwood JPet al. Isogenic P-fimbrial deletionmutants of pyelonephritogenic Escherichia coli: the role ofaGal(14)bGalbinding in virulence of a wild-type strain.Mol Microbiol1993;10: 143155.

47. Roberts JA, Kaack B, Kallenius G et al. Receptors for pyelonephritogenicEscherichia coli in primates. J Urol1984; 131: 163168.

48. Lanne B, Olsson BM, Jovall PAet al. Glycoconjugate receptors for

P-fimbriated Escherichia coliin the mouse. J Biol Chem 1995; 270:90179025.49. Bergsten G, Wullt B, Svanborg C. Escherichia coli, fimbriae, bacterial

persistence and host response induction in the human urinary tract.Int J Med Microbiol2005; 295: 487502.

50. Fischer H, Ellstrom P, Ekstrom Ket al. Ceramide as a TLR4 agonist; aputative signalling intermediate between sphingolipid receptors formicrobial ligands and TLR4.Cell Microbiol2007.

51. Wu XR, Sun TT, Medina JJ.In vitrobinding of type 1-fimbriated Escherichiacolito uroplakins Ia and Ib: relation to urinary tract infections. Proc Natl

Acad Sci USA1996; 93: 96309635.52. Parkkinen J, Finne J, Achtman M et al. Escherichia coli strains binding

neuraminyl a23 galactosides.Biochem Biophys Res Commun 1983; 111:456461.

53. Parkkinen J, Rogers GN, Korhonen T et al. Identification of the O-linkedsialyloligosaccharides of glycophorin A as the erythrocyte receptors forS-fimbriated Escherichia coli. Infect Immun 1986; 54: 3742.

54. Backhed F, Alsen B, Roche N et al. Identification of target tissueglycosphingolipid receptors for uropathogenic, F1C-fimbriatedEscherichia coliand its role in mucosal inflammation. J Biol Chem 2002;277: 1819818205.

55. Khan AS, Kniep B, Oelschlaeger TAet al. Receptor structure for F1Cfimbriae of uropathogenic Escherichia coli. Infect Immun 2000; 68:35413547.

56. Nowicki B, Hart A, Coyne KEet al. Short consensus repeat-3 domain ofrecombinant decay-accelerating factor is recognized by Escherichia colirecombinant Dr adhesin in a model of a cell-cell interaction. J Exp Med1993; 178: 21152121.

57. Virkola R, Westerlund B, Holthofer H et al. Binding characteristics ofEscherichia coliadhesins in human urinary bladder. Infect Immun 1988;56: 26152622.

58. Nowicki B, Rhen M, Vaisanen-Rhen Vet al. Immunofluorescence study offimbrial phase variation in Escherichia coliKS71. J Bacteriol1984; 160:691695.

59. Holden NJ, Gally DL. Switches, cross-talk and memory inEscherichia coliadherence. J Med Microbiol2004; 53: 585593.60. Snyder JA, Haugen BJ, Lockatell CVet al. Coordinate expression of

fimbriae in uropathogenic Escherichia coli. Infect Immun 2005;73:75887596.



Documents

2007-Role of P-Fimbrial-mediated Adherence In