«(Über die Bedeutung der bakteriellen Genomplastizität für die Adaptation und Evolution asymptomatischer Bakteriurie (ABU) Escherichia coli ...»
transcript levels of the gene frmA encoding the glutathione-dependent alcohol dehydrogenase (AdhC) were analyzed in a larger number of in vivo re-isolates (Fig. 39 A). It turned out that frmA expression is up-regulated in other in vivo re-isolates as well. Moreover, frmA expression levels in strains isolated from the same patient were comparable, while they differed from patient to patient. The highest frmA transcript level could be observed in reisolates from patient SR (120-fold higher than in the parent strain). In re-isolates from patients KA and POS, the up-regulation was on average 40-fold and 5-fold, respectively. It is important to mention, that the patient SR was colonized twice during the course of study (Table 11) and that re-isolate SR12 was obtained from an independent inoculation event compared to strains SR3 and SR6. Taken together, the significant up-regulation of the gene frmA could be observed in all re-isolates obtained from the patients KA, SR and POS, indicating that the adaptation of bacterial gene regulation might be patient-specific.
Fig. 39: Real Time-PCR-based quantification of transcript levels of selected genes in ABU re-isolates.
Relative expression of hmpA, metR, tar, iutA and yeiC genes A) and of frmA B) in in-vitro re-isolate
4.9 and all in vivo re-isolates. In all cases, the gene expression of the re-isolates was normalized to that of parent strain 83972. All experiments were performed in triplicate. Gene expression was standardized using the rrnB gene as an internal control.
Taken together, to evaluate changes on the transcriptional level, microarray experiments were performed with three in vivo (CK12, SR12 and KA25) and one in vitro (4.9) re-isolates and compared to that of the parent strain 83972. Genes affected as determined by microarray hybridisation demonstrated remarkable differences among the investigated re-isolates. Each re-isolate represented a unique gene expression pattern, however, a small fraction of genes that were commonly expressed in in vivo re-isolates SR12 and KA25 was detected. Among hypothetical proteins and phage-related genes, most of the changes in the global gene regulation pattern of the in vitro re-isolate 4.9 were implicated in anaerobic respiration. TCA cycle, differed sugars and amino acids transport and metabolism were found to be upregulated in re-isolate SR12 in response to prolonged growth in the bladder. Genes encoding for hypothetical, unclassified and unknown proteins as well phage and transposone related were the main fraction of those individually affected in strains CK12 and KA25. Whereas genes involved in protection against NO (frmAB, hmpA) released upon bacterial infection were commonly up-regalated in re-isolates SR12 and KA25 when compared to parent strain
83972. Moreover, the significant up-regulation of the gene frmA could be observed in all reisolates obtained from the patients KA, SR and POS. Finally, many genes encoding proteins involved in glycine, serine, threonine and methionine transport and metabolism were found to be up-regulated in re-isolates SR12 and KA25 relative to parent strain 83972.
5.5. Changes in the cytoplasmic protein expression of the 83972 re-isolates To analyze changes in protein expression of the re-isolates derived from the human colonization study, a 2D protein gel electrophoresis approach was used. As already described (section 4.3.), cytoplasmic proteins were extracted from bacteria grown in vitro at 37 °C in pooled human urine. The extracted proteins were separated on the basis of their isoelectric point (pH range 4 to 7), followed by a separation according to their molecular weight.
Representative cytoplasmic protein profiles are shown in Fig. 40.
Fig. 40: Comparison of 2D cytoplasmic protein profiles from ABU strain 83972 and the in vivo reisolates KA25, CK12 and SR12 upon growth in vitro at 37 °C in pooled human urine.
5.5.1. Cytoplasmic proteome changes of in vivo re-isolate KA25 relative to parent strain 83972 Altogether, 18 differently expressed proteins were identified (Fig. 41). In accordance with the gene expression profiles on the transcriptional level (section 5.4.1), the number of repressed proteins was higher than that of induced proteins (13 – down, 5 – up) when compared to the cytoplasmic proteome pattern of ancestor strain 83972. These 18 proteins were identified by MALDI-TOF Mass Spectrometry (Table 14). Interestingly, the most striking differences between strains KA25 and 83972 were observed for the proteins FrmA and FrmB which were detectable in higher amounts in re-isolate KA25 than in strain 83972 (Fig. 41). As already described (section 5.4.3), expression of the corresponding genes was strongly up-regulated in strains KA25 and SR12 relative to their parent strain 83972.
Fig. 41: Comparison of the cytoplasmic proteome of ABU strain 83972 and the in vivo re-isolate KA25. Red-channel, cytoplasmic proteins of re-isolate KA25; Green-channel, cytoplasmic proteins of strain 83972. Proteins with the same expression level are shown in yellow. On each gel 300 µg of cytoplasmic proteins were separated and Coomassie-stained. Differently expressed proteins were identified by MALDI-TOF mass spectrometry.
The isoelectric point and molecular weight of the FrmA is very close to those of SerA and GdhA (Table 14). The high amount of FrmA protein might be a reason for problems with GdhA identification in the re-isolate, where spot positions most likely overlap each other. The proteins YdfG and GlyA are involved in serine metabolism while FrmA and FrmB contribute to glutathione metabolism. GdhA is glutamate dehydrogenase that catalyzes the NADPHdependent amination of alpha-ketoglutarate to L-glutamate. It has been reported that impaired metabolism of these amino acids might be due to oxidative and nitrosative stress (Jarboe et al., 2008; Liu et al., 2001). Another over-expressed protein, YtfE, is involved in the repair of damaged iron-sulfur clusters, again due to oxidative and nitrosative stress (Justino et al., 2007).
The GlgB and GlgC proteins, involved in glycogen biosynthesis, were less abundant in the reisolate. In addition, multiple proteins like XylA, AraA, Zwf, PykF which play an important role in the central metabolism were also not as much expressed as in the strain 83972.
Interestingly, two proteins, YeiC and YeiN, which are up-regulated in human urine (Roos et al., 2006b; Snyder et al., 2004), were lesser expressed in re-isolate KA25. YeiC is a putative
sugar kinase from the ribokinase protein family catalysing degradation of pentose sugars. The second protein, YeiN, is not characterized yet. However, it shows more than 90 % homology to the indigoidine synthase A (IndA)-like protein from Thermotoga maritima. Indigoidine is a blue pigment that has been initially described in Erwinia chrysanthemi to be implicated in pathogenicity and protection from oxidative stress (Reverchon et al., 2002). HahA catalyzes the dehydroxylation of bile acids (Yoshimoto et al., 1991). By dehydroxylation, bile acids lose their detergent properties. An increased amount of HahA in re-isolate KA25 might have a positive effect when bacteria grow in the human bladder.
5.5.2. Cytoplasmic proteome changes of in vivo re-isolate SR12 relative to parent strain 83972 The main alterations of the SR12 cytoplasmic proteome relative to that of parent strain 83972 were very similar to that of strain KA25 (Fig. 42).
Fig. 42: Comparison of the cytoplasmic proteome of ABU strain 83972 and the in vivo re-isolate SR12. Red-channel, cytoplasmic proteins of re-isolate SR12; Green-channel, cytoplasmic proteins of strain 83972. Proteins with the same expression level are shown in yellow. On each gel 300 µg of cytoplasmic proteins were separated and Coomassie-stained. Differently expressed proteins were identified by MALDI-TOF mass spectrometry.
Several de-regulated proteins were identified by MALDI-TOF (Table 15). Interestingly, the most striking differences of the SR12 cytoplasmic proteome profiles were exactly the same as those described in re-isolate KA25. The proteins FrmA and FrmB were significantly higher expressed than in the ancestor strain 83972 (Fig. 42). However, the expression level for both proteins in the re-isolate SR12 was twice as much as in KA25 and 100-fold more than both in CK12 and 83972 (Fig. 43).
Table 15: Differently expressed cytoplasmic proteins in the in vivo re-isolate SR12 and the ancestor strain 83972.
Moreover, the YtfE protein involved in NO signalling as well as the proteins SerA, GlyA, GlgB, GlgC, PykF and Zwf were similarly de-regulated as in strain KA25. Nevertheless, YeiC and YeiN were induced in re-isolate SR12 compared to that of re-isolate KA25.
5.5.3. Cytoplasmic proteome changes of in vivo re-isolate CK12 relative to parent strain 83972 The direct comparison of the intracellular proteomes of strain 83972 and re-isolate CK12 allowed the identification of several differently expressed proteins (Fig. 44). In total, eleven candidate spots were detected to differ from the protein profile of strain 83972 (Table 16).
Only one protein which was exclusively found in the CK12 isolate was unidentifiable.
Fig. 44: Comparison of the cytoplasmic proteome of ABU strain 83972 and the in vivo re-isolate CK12. Red-channel, cytoplasmic proteins of re-isolate CK12; Green-channel, cytoplasmic proteins of strain 83972. Proteins with the same expression level are shown in yellow. On each gel 300 µg of cytoplasmic proteins were separated and Coomassie-stained. Differently expressed proteins were identified by MALDI-TOF mass spectrometry.
Whereas the above mentioned proteomes of the re-isolates KA25 and SR12 were very similar to each other, the one of strain CK12 was more distinct. The most prominent features of KA25 and SR12, namely increased amounts of FrmA and FrmB, could not be detected in strain CK12. Also common patterns in expression of YtfE, SerA, GlyA, GlgB, GlgC, XylA,
PykF were not detectable in strain CK12. This might indicate that expression of genes encoding for these proteins is functionally dependent and co-regulated with the frmRAB operon by the same factors. Two proteins, YeiC and YeiN, were identified as commonly deregulated in strains CK12 and KA25. Moreover, only Zwf turned out to be lesser expressed in all three in vivo re-isolates. One protein (Protein #1) which was exclusively expressed in reisolate CK12 but not in the parent strain and other re-isolates could not be identified.
Fig. 45: Adaptation of the ribonucleoside degradation pathway in in vivo re-isolate CK12 of ABU strain 83972. The different steps of ribonucleoside uptake and degradation identified to be deregulated in the different re-isolates are indicated: significantly up-regulated genes according to microarray experiments a); over-expressed proteins according to cytoplasmic proteome comparison b); over-expressed proteins according to outer membrane proteome comparison c).
According to the proteome comparison, only DeoC was found to be present in higher amounts, whereas microarray data indicated that the entire deoCABD operon was upregulated. In addition, transcriptome comparison indicated that the tsx gene encoding for a protein involved in the transport of ribo- and deoxy-nucleosides across the outer membrane of E. coli was up-regulated in re-isolate CK12. This result was also confirmed by comparison of the outer membrane protein expression (section 5.6).
Fig. 46: Adaptation of the deoxy-ribonucleoside degradation pathway in in vivo re-isolate CK12 of ABU strain 83972. The different steps of deoxy-ribonucleoside uptake and degradation identified to be de-regulated in the different re-isolates are indicated: significantly up-regulated genes according to microarray experiments a); over-expressed proteins according to cytoplasmic proteome comparison b); over-expressed proteins according to outer membrane proteome comparison c).
Taken together, to analyze changes in cytoplasmic proteins expression of the re-isolates derived from the human colonization study, a 2D protein gel electrophoresis approach was applied. In all three re-isolates (KA25, SR12, CK12) a number of 18, 12 and 11 proteins, respectively, turned out to be de-regulated. Interestingly, cytoplasmic proteome profiles of isolates KA25 and SR12 were comparable. The most striking differences between those strains and 83972 were observed for the proteins FrmA and FrmB. These results are in a strict accordance with microarray data. The proteome of strain CK12 was more distinct. The most prominent features of KA25 and SR12, namely increased amounts of FrmA and FrmB, could not be detected in strain CK12. Instead, degradation of deoxy- and ribonucleosides was upregulated in strain CK12 compared to SR12, KA25 and 83972.
5.6. Outer membrane proteome changes of the in vivo re-isolates of strain 83972.
Bacterial surface proteins have different main functions. They are involved in the uptake and transport of different substances into the cell, in adhesion, and in cell-to-cell communication.
Microorganisms growing in the urinary tract encounter iron depletion. Thus they must have evolved multiple iron uptake mechanisms. Moreover, all proteins located on the bacterial surface are potential targets for antibodies of the human immune system.
To extend our knowledge about changes in the outer membrane proteome and adaptation of ABU strain 83972 during growth in the human bladder, the membrane protein fractions of this strains and its in vivo re-isolates were compared. As already described (section 4.3), following bacterial growth in pooled human urine, carbonate-insoluble proteins were separated by 2D gel electrophoresis. Protein spots differing between the parent strain 83972 and its derivatives were identified by MALDI-TOF mass spectrometry.