Type IV pili are expressed on the surface of gram-negative bacteria and are associated in a range of functions that mediate pathogenesis. PilP is one component of the molecular machine of proteins that work in concert to co-ordinate the assimilation type IV pili. To characterize PilP and determine its ligand binding properties a folded domain of PilP (PilP(d)) was transformed into Escherichia coli and the protein was overexpressed and purified to homogeneity by a combination of metal affinity and size exclusion chromatography and dialysis. Purified PilP(d) exhibited low-affinity binding to the fluorescence hydrophobic probe ANS, displaying an apparent dissociation constant of 14.07 ïM, and no affinity for Nile Red. Fluorophores were found to have an intrinsic fluorescence increase over time with no protein present thereby masking the typical change in emission spectra observed on complex formation. The intrinsic increase in fluorescence was not due to a photobleaching affect, although the solvent polarity had an effect which could account for the around 15 % of the intrinsic increase observed. Taking all this into consideration it can be suggested that the intrinsic increase in fluorescence displayed by ANS is so much so, that ANS binding to PilP(d) is masked. Therefore any quench in fluorescence when competitive ligands are titrated in to displace ANS cannot necessarily be attributed to an efficacious displacement. Alternative twitching motility screening assays were used to screen inhibitor ligands and show a better indication of ligand efficacy, although which component of the type IV biogenesis molecular machinery is being effected cannot be inferred. Three ligands were found to have an effect on twitching motility, 5 - Amino -1,2,3,4 - tetrahydroisoquinoline, 5-Aminosioquinoline and 4-Aminochinaldin, without inhibiting growth concurrently.
1.0 INTRODUCTION
Type IV pili play a pivotal role in bacterial pathogenesis. Extension and retraction of polar filaments generates a powerful motorforce enabling bacteria to adhere too and colonise certain surfaces during infection. In Pseudomonas and Neiserria species characteristic surface appendages are also implicated in twitching motililty and biofilm formation (1). The process is co-ordinated by a team of proteins that work in concert to extrude polar pili subunits over the selectively permeable bacterial membrane into the extracellular space. After adhesion and interactions with the target cell and/or environment the bacteria are capable of generating ATP dependant retractile forces to disassemble and pull the filaments back in (2). Upto 40 different protein players have been identified to date, some of which are homologous to those of the general (type II) secretion pathway, that rely on spatially and temporally co-ordinated protein-protein interactions to mediate pathogenesis.
Using small ligand inhibitors that have been selected by structure based rational drug design it may be possible to prevent critical protein-protein interactions. Blocking discrete steps will allow insight into how protein components interact with each other to assemble the molecular machine, a process which until now has remained elusive.
PilP - A Component of the Type IV Pilus Biogenesis Molecular Machine
The lipoprotein PilP is one of around 15 different proteins have been identified to play an integral role in molecular machinery assembly. The 12 kDa protein copurifies with the inner membrane fraction and has been implicated in the stabilisation of outer membrane protein PilQ in Neisseria (3). The association of these two proteins sees the formation of a pore in the outer membrane through which pili subunits are extruded during pilus biogenesis (4). Gene knockout studies have also shown that PilP is required for pilus biogenesis, with pilus expression and transformation being abolished in mutants (5).
Figure 1.0 Ribbon Representation of PilP(69-181) folded domain. Purple arrow used to indicate the hydrophobic putative binding cavity with conserved amino acid residues. Seven Β-strands depicted in green, and one α-helix in orange, unstructured N and C terminal shown with a black line.Golovanov et al solved the structural of a folded domain from the Neisseria Meningitidis PilP protein (Figure 1.0). "Residues 85-163 of the domain adopt a β-sandwich-type fold; a three stranded up-and-down antiparallel β-sheet is packed against an up-and-down four stranded antiparallel β-sheet. A short helix is formed at the N terminus and makes hydrophobic interaction with residues in the β-sheets. A crevice lined with hydrophobic residues is identifiable between the two β-sheets packed together" (6, 7). The structure predicts a region of flexibility and disorder from residues 1-85 so the crystal structure was solved with a recombinant version of PilP with these residues removed. Subsequently the structure of PilP from Pseudomonas aeruginosa has been solved and is consistent with this crystal structure (60). It has been predicted that the hydrophobic crevice could be a suitable site for a small inhibitory ligand, with the intended aim of preventing protein-protein interaction that has a knock-on effect on type IV pilus biogenesis. However it remains to be shown whether this binding site is implicated in interactions with PilQ as another study has shown interaction occurs specifically through the N-terminal and C-terminal segments of PilP with the central portion of PilQ (3).
Probing PilP Using Structure Based Rational Drug Design -
The 3D crystal structure of PilP's folded domain (hereby referred to as PilP(d)) and identification of a putative binding site suitable for a small ligand inhibitor opens up the field of structure based rational drug design. Traditional High-throughput screening techniques have been ruled out to characterise PilP(d) due to the massive computational and man power required for library screening which are not practical on this scale. Using ligand-based and structure-based design it is possible to narrow the search down to screen a more manageable library of compounds. This allows for careful experimental assays on a small number of database compounds based on their predicted suitability for the active site (8). Bioinformatic and chemoinformatic technologies can be used in combination with the three dimensional structure of PilP(d) to rationally design and virtually screen small compound libraries. Using "Docking" programs it is possible to take the lock and key concept of protein-ligand specificity to assess the geometric complimentarity of suitable ligands. This allows for ligand optimisation in the context of a binding site, the best compounds then being tested experimentally for active site affinity and activity in appropriate assays. This method has proved successful in the identification of an inhibitor to prevent the development of infectious larva from Brugia pahangi by inhibiting the essential cathepsin L-like cysteine protease (9).
Using inhibitors selected through a combination of docking methods and rational geometry optimisation we took two completely different approaches to further functionally analyse the putative binding site of PilP, in vitro and in vivo.
In vitro -
Fluorescence experiments were used to probe PilP(d) to evaluate it's affinity for potential novel ligand inhibitors. The principle behind this being that when ligands were titrated into the cuvette, bound fluorophores would be displaced from the hydrophobic crevice creating a detectable quench in fluorescence. In order to quantify the effect of novel ligand inhibitors it was a prerequisite to determine the binding of fluorophores to PilP(d).
The use of fluorophores (Figure 1.1) is well documented and a popular tool for protein characterisation. Their versatility and sensitivity make them suitable for a range of applications in determining protein interactions. In conjunction with steady state fluorescence spectroscopy they have proved to be a useful in a variety of assays. The early literature demonstrates their range of uses from detecting the molten globule interactions (10) to probing the active site of enzymes (11) and also as a tool to assess the surface hydrophobicity (12). But more recently they are still being used as a tool for protein characterisation and have been demonstrated as assays that monitor unfolding and refolding processes (13) or probe protein surfactant interactions (14) plus many more.
The Effect of Fluorophore Microenvironment on Fluorescence Spectra
The mechanisms of fluorophore action can be attributed to their intrinsic properties, and in this circumstance, the presence of an extended π system. Excitation by absorbtion of light lifts the electrons of the dye molecules within femotoseconds from the ground state to higher singlet excited states (15). Subsequent loss of vibrational energy, and reversion to the lower energy states again results in a detectable emission of a photon at a different wavelength to excitation. Electronic excitation of aromatic compounds typically results in an increase in dipole moment; thus excited molecules exhibit larger dipole moments than molecules in the ground state (16). The surrounding solvent responds to the increased dipole moment by reorganising it's molecules around the fluorophore - a concept known as solvent relaxation. A red shift in emission spectra (Stokes Shift) results due to the efficiency of vibrational relaxation through which energy is lost to the solvent and a lower fluorescence emission is observed (17). These intrinsic properties and the tendancy of fluorophores to fluorescence according to their microenvironment are the principles that we intended to exploit in order to characterise the nature of PilP's binding site.
The effect of microenvironment on fluorescence spectra is demonstrated using the extended π system in tryptophan as an intrinsic fluorophore. When tryptophan residues are buried in the hydrophobic core of a protein, their emission spectra reflects the local environment compared to those that are on the surface of the protein. As the conformation of the protein changes, and the tryptophan becomes more exposed, the solvent dependant emission spectra shifts to a shorter wavelength and the intensity of the fluorescence increases (18). The switch from noncovalent interactions with a protein to interaction with the polar solvent exposed environment causes measurable changes in fluorescence spectrum. This can be detected by either changes in emission intensity or as a shift in maximum emission, or both (19). This tool has proved to be a particularly powerful due to the finding that tryptophan can often be substituted for other amino acids by site-directed mutagenesis, with minimal effect on structure and activity (20).The conformation of Vibrio harveryi acyl carrier protein has been monitored using this technique.
There are numerous studies in the literature describing a similar concept to characterise hydrophobic protein binding sites, the difference being the use of extrinsic dyes such as ANS and Nile Red (see figure 1.1 for structures). Brown et al utilised the fluorescence properties of both ANS and Nile red in tandem to probe the binding sites of Human Serum Albumin (HSA) (21). Due to their disparities in emission wavelength and preference for sites of differential polarity on protein surfaces they showed not only the presence of multiple binding sites on HSA but also the nature of binding sites according to fluorophore affinites. Both fluorophores exhibited a higher fluorescence intensity when docked into their respective binding sites thereby shielding them from the external environment, demonstrating that the fluorescence intensity is directly dependant on the surrounding microenvironment (21).
Considering the hydrophobic binding crevice of PilP, and the tendancy of fluorophores to bind these local environments, it is reasonable to hypothesise that this protein would be suitable candidate for similar fluorophore binding experiments.
Figure 1.1 Fluorophores ANS (A) and Nile Red (B). Extended π system gives them a detectable fluorescence depending on solvent polarity. A property that is commonly exploited for protein characterisation
(B)
(A))Assuming that fluorophore binds the aforementioned hydrophobic cavity, this opens up the potential to study novel ligand inhibtors designed to displace ANS. Kane et al were able to show the preference of keratinocyte-lipid binding protein (KBLP) for binding long chain fatty acids over shorter chain fatty acids based upon displacement of bound fluorophore (22). The same princinple applies here; theoretically ANS will bind to the hydrophobic cavity of PilP, showing an increase in emission intensity accompanied by a shift in maximum emission. An effect which will be quenched when ligands with higher affinity compete and displace bound ANS.
Practical Considerations for Fluorescence Experiments
There are a number of practical issues that need to be considered for fluorescence experiments; here the issues relating to this study will be mentioned.
Firstly the complex photophysical properties of fluorophores need to be considered. Due to the two differentially excited states which arise according to it's surrounding polarity it must be noted that it is possible to effect the quantum yield and emission energy spectra according to solvent polarity (23-25). The solvent molecules surrounding the fluorophore are able to re-arrange to more energetically favorable positions prior to emission, effectively lowering the energy of the excited state. In practical terms the solvent environment of ANS has to be kept stable in order to mininize the effect on fluorescence intensity in different conditions.
Secondly photobleaching can occur due to the successive exposure of sample to high-intensity light beam characterised by a decrease in sample fluorescence (26). A study looking at the "bleaching effect" on nine different proteins showed that with fast fluorescent readings at a fixed wavelength, followed by mathematical interpolation of the data this effect can be restricted, which can account for upto 15% of the signal in some cases (27). Care must be taken to reduce the exposure of fluorophore to irradiation.
Thirdly the inner filter effects (IFE) need to be considered. A decrease in emission intensity or distortion of bandshape may be observed as a result of reabsorbtion of emitted radiation. Similarly the absorption of incident radiation by a species other than the intended primary absorber may also influence the emission spectra. Carrying out the assays in vitro minimizes the interaction of the fluorophore with anything but the protein and therefore prevents the IFE distorting the emission spectra. Diluting the sample to an acceptable absorbance level has also been shown to reduce the inner filter effects, although this may introduce increased contamination from diluting water or cause perturbations in the stability of the protein (28)
In Vivo
In vivo assays were designed to screen for the inhibition of bacterial twitching motility, a process that is dependant on PilP. The theory being that if novel ligands have a positive association with PilP, the assimilation of type IV pili would be inhibited, thereby preventing twitching motility.
As well as their role in biofilm formation (29), virulence (30, 31) and prokaryotic horizontal gene transfer (32-34), type IV pili are also integral to twitching (1, 35). Twitching motility refers to the flagella-independent form of bacterial translocation over moist surfaces through extension, tethering and retraction of type IV pili (36). It appears as a means for bacteria to travel in environments with lower water contents and allows colonization of hydrated surfaces (37). It is a known virulence factor, causing spreading on body surfaces during infection (1, 38)
Twitching motility can visualised in wild-type Pseudomonas aeruginosa as flat, spreading colonies with characteristic rough appearance, distinguishable from growth by a small peripheral twitching zone consisting of a thin layer of cells (37). Active twitching is also exhibited at the interstitial surface between agar or other nutrient gel and plastic or glass, a characteristic which can be exploited and used as a simple assays for assessing twitching phenotypes as shown by Chiang P et al (39). Inoculation at a single point incites radial colonial expansion, with different bacterial phase variants recovered in each zone (40). Twitching motility can be attributed to the fine outmost ring or "halo" of single cells, which is completely absent in twitching deficient mutants (41). These phenotypes also show colonisation on the surface of the agar indicating normal proliferation has not also been affected.
A study by JC McMicheal went on to further characterise the individual cells found in the circular colonies that had been visualised through staining with coomassie blue. They found that individual adherent bacteria in the outer rings spontaneously moved short distances, whereas cells in the inner ring had entered a quiescent state (40). They predicted that this phenotype could be explained by the presence of pili structures in the outermost ring, where retraction after tethering to surfaces at their distal end allowed cells to move forward and colonize new surfaces. Indeed, blotting onto nitrocellulose and probing with anti-pili specific antibodies confirmed the presence of pili in the outer ring. It is now commonly accepted that the observable rapid expansion by these cells is dependant on pilus associated twitching motility (42, 43)
Because twitching motility has absolute dependence on type IV pili, which is dependant on PilP, this serves as a useful indicator to demonstrate where ligands may have bound to PilP. Although we cannot say for certainty that type IV pilus biogenesis is prevented through the inhibition of PilP, it may be inhibiting other proteins of the pilus biogenesis molecular machinery. The primary purpose is a screening assay to narrow down the search for ligands with inhibitory properties, ligands that may have been previously overlooked in any conventional antibiotic screen because they have no effect on the ability of bacteria to grow. In theory bacterial colonies grown in the presence of inhibitory ligands will have the same phenotype as PilP knockout strains through inhibition at the proteinomic level, as opposed the genomic level. The short-term goal being the identification of ligands that bind to PilP suitable for NMR or co-crystallographic studies. Ultimately such small molecules could then be developed as tools to elucidate the mechanisms of bacterial pathogenesis, preventing protein-protein interactions at each stage of nanomachine assembly. In the long term optimisation of the small lead compounds could be developed for therapeutic purposes agents for the prevention or treatment of bacterial colonisation (44). Targeting virulence control through inhibiting type IV pili biogenesis is an attractive target for disease management and a useful alternative in the fight against the antibiotic resistant bacteria.
Aims -
Use intrinsic fluorescence properties of fluorophores ANS and Nile Red to probe the putative hydrophobic binding site of PilP(d), in vitro, using fluorescence spectroscopy. If ANS binds to the crevice of PilP(d) it should produce a fluorescence signal
To screen ligands for their ability to displace ANS through fluorescence quenching
Screen ligands In vivo for their ability to inhibit twitching motility.
4.0 Discussion
A domain of His-Tagged PilP (PilP(d)) was overexpressed in Escherichia coli and purified. The protein purified was assumed to be PilP(d) due to its molecular weight and the presence of a His-tag. Previous western blot assays with antibodies raised against PilP(d) have confirmed the presence of PilP(d) after the purification process. The transcript of PilP that was overexpressed had been designed with N terminal residues 1-82 cleaved. This removed a flexibility region and a prokaryotic membrane attachment site that is predicted to be unstructured in the crystal. The protein therefore remained in the soluble fraction, which would otherwise co-purify with the inner membrane fraction (3). It may be possible that this terminus is integral to interaction with PilQ during pilus biogenesis as shown by Balasingham et al 2007. However Golovonav et al have shown that the folded portion of PilP(d), which has the small hydrophobic crevice we are interested in characterizing, is unaffected by the removal of N-terminal residues (6). Therefore the specificity and functionality of this region is likely to have been retained in PilP(d). Furthermore it is typically assumed that protein-protein interactions occur through folded conserved domains and it is on this assumption that we went on to probe the putative binding site.
Fluorescence experiments were carried out in order to explore the binding properties of the hydrophobic binding crevice on PilP(d). Before displacement assays with competitive ligands could be carried out, it was a necessary to assess the affinity of fluorophores ANS and Nile Red for PilP(d) (45). We predicted that the hydrophobic crevice of PilP(d) would provide a suitable binding site for ANS and Nile Red thereby shielding their extended π system from the local solvent. Using spectrofluorimetry the resulting change in fluorescence would be detected. By scanning the emission spectra at a set excitation wavelength it was possible to show the shift in fluorescence intensity and blue shift in emission maximum that is characteristic of protein-fluorophore interaction spectra. The in vitro fluorescence experiments show that compared to BSA, fluorophores ANS and Nile Red bind to PilP(d) with low affinity and no affinity respectively. There was also no typical blue shift in emission maximum that is usually observed as a result of a protein-fluorophore interaction. However as shown by Brown et al this may only be detectable at high binding affinities, as seen with BSA. For ANS it is likely that protein doesn't completely shield the fluorophore from its local environment so only a small change in fluorescence is observed. Whereas the results from Nile Red indicate that no complex is formed with ANS and the extended π system remains exposed to the surrounding solvent .
The number and nature of protein binding sites can also account for a change in emission spectra. Using similar fluorescence experiments with BSA another group found that there were four ANS binding sites detectable through spectrofluorimetry. With increased binding sites for ANS a much higher proportion of ANS is shielded from the polar environment, which can be directly observed by a greater fluorescence intensity. This explains the observed high increase in fluorescence intensity compared to PilP(d). The lower increase in emission intensity we observed for ANS binding to PilP(d) being put down to fewer binding sites.
Furthermore, other groups have reported even more binding sites for ANS on BSA (46, 47), identified through equilibrium and potentiometry, not detectable by fluorescence measurement due to their more hydrophillic nature. This may also be true for PilP(d) and another reason contributing to the low fluorescence emission intensity detected. As well as having fewer sites for ANS, if the crevice is not hydrophobic enough, water molecules can gain access to the bound ANS ions to quench their fluorescence (48). Although it must be noted that there is a large range of reported binding constants and binding sites per protein in the literature for BSA/ANS, it is assumed there are more binding sites than on PilP(d)
We found that ANS had an unexplainable intrinsic increase in fluorescence over time. This observation was even apparent with no protein present so cannot be put down to an extremely slow folding protein, although the increase did plateau when the cell was left for 2 hours. It was possible to minimize the effect using H2O as a fluorophore solvent, instead of more polar ethanol. As predicted a change in fluorescence intensity was observed according to the solvent polarity. Previous work by other groups have shown that as the hydrogen bonding potential of a solvent increases, the quantum yield of fluorescence decreases (16). The large fluorescence intensity observed with ethanol compared to water as the solvent can be explained due to its inability to form hydrogen bonds with ANS. The greater polarity of water allows re-distribution of electrons according to the altered dipole moment of the excited fluorophore. In terms of fluorescence intensity this can be seen as a decrease in quantum yield, as our experiments demonstrate.
The low increase in fluorescence intensity we observed indicates that ANS binds to PilP(d) with low affinity. Therefore the intrinsic increase in fluorescence may have been so much so, that it masked any small sensitive changes in fluorescence spectra arising from PilP(d) -ANS complex formation. In the case of BSA binding to ANS the large increase in fluorescence intensity makes the intrinsic increase relatively insignificant and can therefore be disregarded when characterizing protein-ligand interaction in this instance.
We postulated that the intrinsic increase in fluorescence over time observed by ANS could be due to a photobleaching effect as a result of from long exposure to exciting radiation (49). Typically limiting the exposure to excitation can solve this effect, or reducing the intensity of exciting light, but it has been shown that this decreases the signal to noise ratio and compromises the data quality. This is of particular importance with respect to PilP(d) due to the nature of its binding and the sensitivity needed to detect small changes in the fluorescence spectra. To rule out the possibility that the intrinsic increase in fluorescence was a result of photobleaching the sample was exposed to increasing excitation radiation, which was found to have no effect on fluorescence intensity. This is consistent with a group who found that they were able to minimize the photobleaching effect to 2.4 % in Blue Fluorescent Protein ( BFP) using a Varian Cary Eclipse Spectrophotometer due to the xenon flashlamp that irradiating the sample only when readings are being taken (49)(50).
Fluorophore displacement reactions were carried out with ligands identified through structure-based rational design (51). Ligands were selected according to their predicted affinity for the hydrophobic binding crevice identified in the crystal structure (60). A similar assay has been done to characterize and determine the ligand binding properties of keratinocyte lipid-binding protein (KBLP) where preliminary X-ray crystallographic analysis has shown that ANS resides within the binding cavity, as opposed to our rational prediction. Competition assays were carried out and various lipid-ligands were identified with the ability to displace ANS bound to KLBP (22). Our assays revealed that no ligands displayed an apparent ability to displace ANS as no fluorescence quenching was observed. Again the intrinsic increase in fluorescence must be kept in mind when considering the change in fluorescence intensity. It is possible that the detectable decrease in % of ANS bound to increasing lipid-ligand concentration as observed by Kane et al may have been masked in the case of PilP(d) by the intrinsic increase in ANS fluorescence.
The difficulties we encountered characterizing PilP(d) 's binding site, namely its low affinity for ANS, meant an alternative assay had to be used to screen potential pilus biogenesis small ligand inhibitors. From the fluorescence experiments it was not possible to definitively conclude that the ligands had no affinity for the binding site of PilP(d) so in vivo assays were implemented. They were designed in order to approach the mechanisms of pilus biogenesis from a completely different angle. Because pilus biogenesis is integral to twitching motility (4), strains which have had their pilus biogenesis machineries inhibited can be characterized by their inability to twitch whilst retaining normal colonization on the agar surface.
We identified three ligands that had an apparent ability to inhibit twitching motility without inhibiting growth concurrently, i.e there was no characteristic "halo" of cells, but maintained growth on the surface of the agar. It cannot be definitively concluded that twitching motility is inhibited through interaction with PilP however, or indeed through direct inhibition of the type IV pilus system. The type IV pilus structure located within the bacterial cell wall is a sophisticated multimeric protein machine. The events that lead to type IV pilus biogenesis are highly spatially and temporally regulated involving the integration and coordination of each component of the nanomachine (52). The ligands could therefore be associating with any one of up to 40 components to prevent twitching motility (36). For example PilB, PilC and PilD are protein components from Pseudomonas aeruoginsa that are also considered essential for pilus biogenesis (53). Mutants have a phenotype that is unable to produce surface assembled type IV pili and are non-twitching. Just as inhibitory ligands may be binding to PilP, they may also be exerting their inhibitory effects through these proteins, or indeed through other proteins that are associate with twitching motility, that are not part of the type IV pilus system, and no phenotypic difference would be measurable.
However the ligands we identified do still provide a promising start in the search for twitching motility inhibitors. Interestingly all three ligands share similar structural features. They are polycyclic hydrocarbons consisting of two fused aromatic rings, all with an amine group on the fourth or fifth carbon in the ring. Ligands 5 and 9 also have an integral nitrogen atom contributing to the aromatic rings, which is not present in ligand 13 (see figure 3.3.2 for structures). Due to their structural similarities it is reasonable therefore to postulate that they are exerting their inhibitory effects through the same site, albeit an unknown site. These results reflect similar findings by LM Junker et al. High-throughput screening methods identified 30 compounds that have inhibitory effects on biofilm formation, a process that is also dependent on type IV pili (29). Of the 30 compounds identified they all fell into just 6 different structural classes, with each category sharing similar structural features (44). It would be interesting to see whether the ligands identified by Junker and Clardy have an inhibitory effect on twitching motility, and vice-versa, whether the ligands we identified have an inhibitory effect on biofilm formation. Similarly another group have used an Oroidin library to screen against biofilm formation and noticed that the most active analogues identified were those that contained a 2-aminoimidazole motif and dibrominate pyrrolecarboamide subunit (54, 55).
The presence of a concentrated ring of cells on the agar plates, combined with colonization on the agar surface was used to characterize twitching motility. We used increasing concentrations of ligands but encountered problems quantifying the effect of ligand concentration on inhibition. Because the number of cells used to inoculate the agar was not standardized (i.e. it was unknown) problems arose measuring the twitching inhibition in relation to growth. Some colonies only displayed a small amount of growth on the agar plate, but still retained a clear twitching halo on the plate surface. Through semi-quantitive classification of inhibition we concluded that ligands 2 (2,8 - Quinolinediol),11 (4-Phenylimidazol), 12 (1,8 - Napthalimid), 17 (2 - Chloroquinoxaline), 18 (Trans-2-methyl-3-phenyl-2-propen-1-ol), and 22 (Ethyl 2-chloronicotinate) were toxic to the cells, preventing surface colonization. Although our ability to draw conclusions on the degree of inhibition is limited, as colony size depends on the number of cells that were selected for inoculation. The assays still served their main purpose in that they provided a useful and easily applicable screen to narrow down the search for type IV pili inhibitors.
Further Directions
There are numerous directions that this project could take in order to further characterize PilP-ligand interactions. For example Isothermal Calorimetry - Panse et al have shown the thermodynamics of substrate binding to the Chaperone SecB using a combination of ITC and fluorescence Spectroscopy. A complete thermodynamic profile could confirm the biomolecular interaction between PilP and inhibitory ligands. However the protocols generally reply on a large quantity of protein for accurate measurements. Although the typical issues of protein availability are less of a problem with PilP (as we have shown good yield is achievable from the overexpression and purification steps), it is still desirable to use an assay that minimizes protein consumption. Therefore Surface Plasmon Resonance would be suitable alternative for direct and rapid small ligand library screening and constant determination (56). By immobilizing PilP(d) on a Dextran surface it would be possible to run multiple ligands simultaneously over the chip surface, a more sensitive assay, to detect the SPR signal emitted on biomolecular interaction (57).
Successful identification of a ligand that interacts with the binding site on PilP also leads to further steps in protein characterization. The crystal structure solution of PilP from Pseudomonas can be used to detect ligand binding and to determine ligand orientation. This also gives rise to the possibility of co-crystallographic studies, soaking of protein crystals into inhibitor can be used to confirm where and how ligands are interacting with PilP. The crystal structure of the MurA, an essential enzyme for bacteria wall biosynthesis, has been solved complexed with ANS. Schonbrunn et al show that substrate binding induces a change in conformation that explains the previous fluorescence observations with ANS (58).
Inhibiting protein-protein interactions using small inhibitory compounds serves a useful aid in dissecting the mechanisms of assembly of disassembly of type IV pili. Structural determination of pilus biogenesis proteins will help to shed light on how they are interacting with each other and co-coordinating pathogenesis. Targeting bacterial virulence factors as opposed to typical antibiotics will help to identify new players in the fight against ever-emerging antibiotic resistant strains of bacteria.
1. Henrichsen, J. (1983) Twitching motility., Annu Rev Microbiol 37, 81-93.
2. Satyshur, K., Worzalla, G., Meyer, L., Heiniger, E., Aukema, K., Misic, A., and Forest, K. (2007) Crystal structures of the pilus retraction motor PilT suggest large domain movements and subunit cooperation drive motility., Structure 15, 363-376.
3. Balasingham, S., Collins, R., Assalkhou, R., Homberset, H., Frye, S., Derrick, J., and Tønjum, T. (2007) Interactions between the lipoprotein PilP and the secretin PilQ in Neisseria meningitidis., J Bacteriol 189, 5716-5727.
4. Drake, S., and Koomey, M. (1995) The product of the pilQ gene is essential for the biogenesis of type IV pili in Neisseria gonorrhoeae., Mol Microbiol 18, 975-986.
5. Drake, S., Sandstedt, S., and Koomey, M. (1997) PilP, a pilus biogenesis lipoprotein in Neisseria gonorrhoeae, affects expression of PilQ as a high-molecular-mass multimer., Mol Microbiol 23, 657-668.
6. Golovanov, A., Balasingham, S., Tzitzilonis, C., Goult, B., Lian, L., Homberset, H., Tønjum, T., and Derrick, J. (2006) The solution structure of a domain from the Neisseria meningitidis lipoprotein PilP reveals a new beta-sandwich fold., J Mol Biol 364, 186-195.
7. Golovanov, A., Balasingham, S., Tzitzilonis, C., Goult, B., Lian, L., Homberset, H., Tønjum, T., and Derrick, J. (2006) Assignment of 1H, 13C, and 15N resonances for the PilP pilot protein from Neisseria meningitidis., J Biomol NMR 36 Suppl 1, 68.
8. Lyne, P. (2002) Structure-based virtual screening: an overview., Drug Discov Today 7, 1047-1055.
9. Selzer, P. M., J., M. R., and Andreas, R. (2008) Protein Structures and Structure-Based Rational Drug Design, Springer Berlin Heidelberg.
10. Goto, Y., and Fink, A. (1989) Conformational states of beta-lactamase: molten-globule states at acidic and alkaline pH with high salt., Biochemistry 28, 945-952.
11. Takashi, R., Tonomura, Y., and Morales, M. (1977) 4,4'-Bis (1-anilinonaphthalene 8-sulfonate) (bis-ANS): a new probe of the active site of myosin., Proc Natl Acad Sci U S A 74, 2334-2338.
12. Cardamone, M., and Puri, N. (1992) Spectrofluorimetric assessment of the surface hydrophobicity of proteins., Biochem J 282 ( Pt 2), 589-593.
13. Acharya, P., and Rao, N. (2003) Stability studies on a lipase from Bacillus subtilis in guanidinium chloride., J Protein Chem 22, 51-60.
14. Kumar, V., Sharma, V., and Kalonia, D. (2005) Second derivative tryptophan fluorescence spectroscopy as a tool to characterize partially unfolded intermediates of proteins., Int J Pharm 294, 193-199.
15. Hawe, A., Sutter, M., and Jiskoot, W. (2008) Extrinsic fluorescent dyes as tools for protein characterization., Pharm Res 25, 1487-1499.
16. Barghouti S, A., J, P., and L.H, H. (1998) Effects of Solvents on the Fluorescence Emission Spectra of 1-Anilino-8-Napthalene Sulfonic Acid: A physical Chemistry Experiment, Vol. 3, Springer-Verlag New York Inc.
17. Samanta, A. (2006) Dynamic stokes shift and excitation wavelength dependent fluorescence of dipolar molecules in room temperature ionic liquids., J Phys Chem B 110, 13704-13716.
18. Mocz, G. (2007) Fluorescent proteins and their use in marine biosciences, biotechnology, and proteomics., Mar Biotechnol (NY) 9, 305-328.
19. Moller, M., and A, D. (2002) Biochemistry and Molecular Biology Education, Vol. 5, The Internation Union of Biochemistry and Molecular Biology.
20. Vivian, J., and Callis, P. (2001) Mechanisms of tryptophan fluorescence shifts in proteins., Biophys J 80, 2093-2109.
21. Brown, M. B., E, E. T., M, M. J., and J, S. N. (1993) Use of Nile Red as a long-wavelength fluorophore in dual-probe studies of ligand-protein interactions, Journal of Fluorescence, 129-130.
22. Kane, C., Coe, N., Vanlandingham, B., Krieg, P., and Bernlohr, D. (1996) Expression, purification, and ligand-binding analysis of recombinant keratinocyte lipid-binding protein (MAL-1), an intracellular lipid-binding found overexpressed in neoplastic skin cells., Biochemistry 35, 2894-2900.
23. Kosower, E., Hanna, D., and Hanna, K. (1978) Intramolecular donor-acceptor system. 4. Solvent effects on radiative and nonradiative processes for the charge-transfer states of N-arylaminonaphthalenesulfonates.
24. Gasymov, O., Abduragimov, A., and Glasgow, B. (2007) Characterization of fluorescence of ANS-tear lipocalin complex: evidence for multiple-binding modes., Photochem Photobiol 83, 1405-1414.
25. Gasymov, O., Abduragimov, A., and Glasgow, B. (2007) Evidence for internal and external binding sites on human tear lipocalin., Arch Biochem Biophys 468, 15-21.
26. Lakowicz, J. R. (2006) The Principles of Fluorescence Spectroscopy, Third ed.
27. Di Stasio, E., Bizzarri, P., Misiti, F., Pavoni, E., and Brancaccio, A. (2004) A fast and accurate procedure to collect and analyze unfolding fluorescence signal: the case of dystroglycan domains., Biophys Chem 107, 197-211.
28. Larsson, T., Wedborg, M., and Turner, D. (2007) Correction of inner-filter effect in fluorescence excitation-emission matrix spectrometry using Raman scatter., Anal Chim Acta 583, 357-363.
29. O'Toole, G., and Kolter, R. (1998) Flagellar and twitching motility are necessary for Pseudomonas aeruginosa biofilm development., Mol Microbiol 30, 295-304.
30. Bieber, D., Ramer, S., Wu, C., Murray, W., Tobe, T., Fernandez, R., and Schoolnik, G. (1998) Type IV pili, transient bacterial aggregates, and virulence of enteropathogenic Escherichia coli., Science 280, 2114-2118.
31. Comolli, J., Hauser, A., Waite, L., Whitchurch, C., Mattick, J., and Engel, J. (1999) Pseudomonas aeruginosa gene products PilT and PilU are required for cytotoxicity in vitro and virulence in a mouse model of acute pneumonia., Infect Immun 67, 3625-3630.
32. Bradley, D. (1972) Evidence for the retraction of Pseudomonas aeruginosa RNA phage pili., Biochem Biophys Res Commun 47, 142-149.
33. Seifert, H., Ajioka, R., Marchal, C., Sparling, P., and So, M. (1988) DNA transformation leads to pilin antigenic variation in Neisseria gonorrhoeae., Nature 336, 392-395.
34. Yoshida, T., Kim, S., and Komano, T. (1999) Twelve pil genes are required for biogenesis of the R64 thin pilus., J Bacteriol 181, 2038-2043.
35. Wall, D., and Kaiser, D. (1999) Type IV pili and cell motility., Mol Microbiol 32, 1-10.
36. Mattick, J. (2002) Type IV pili and twitching motility., Annu Rev Microbiol 56, 289-314.
37. Mattick, J., Whitchurch, C., and Alm, R. (1996) The molecular genetics of type-4 fimbriae in Pseudomonas aeruginosa--a review., Gene 179, 147-155.
38. Hahn, H. (1997) The type-4 pilus is the major virulence-associated adhesin of Pseudomonas aeruginosa--a review., Gene 192, 99-108.
39. Chiang, P., and Burrows, L. (2003) Biofilm formation by hyperpiliated mutants of Pseudomonas aeruginosa., J Bacteriol 185, 2374-2378.
40. McMichael, J. (1992) Bacterial differentiation within Moraxella bovis colonies growing at the interface of the agar medium with the Petri dish., J Gen Microbiol 138, 2687-2695.
41. Alm, R., and Mattick, J. (1995) Identification of a gene, pilV, required for type 4 fimbrial biogenesis in Pseudomonas aeruginosa, whose product possesses a pre-pilin-like leader sequence., Mol Microbiol 16, 485-496.
42. Pedersen, K. (1972) Isolation and description of a haemolytic species of Neisseria (N. ovis) from cattle with infectious keratoconjunctivitis., Acta Pathol Microbiol Scand B Microbiol Immunol 80, 135-139.
43. Pedersen, K., Froholm, L., and Bovre, K. (1972) Fimbriation and colony type of Moraxella bovis in relation to conjunctival colonization and development of keratoconjunctivitis in cattle., Acta Pathol Microbiol Scand B Microbiol Immunol 80, 911-918.
44. Junker, L., and Clardy, J. (2007) High-throughput screens for small-molecule inhibitors of Pseudomonas aeruginosa biofilm development., Antimicrob Agents Chemother 51, 3582-3590.
45. Chipman, D., Grisaro, V., and Sharon, N. (1967) The binding of oligosaccharides containing N-acetylglucosamine and N-acetylmuramic acid to lysozyme. The specificity of binding subsites., J Biol Chem 242, 4388-4394.
46. Kolb, D., and Weber, G. (1975) Cooperativity of binding of anilinonaphthalenesulfonate to serum albumin induced by a second ligand., Biochemistry 14, 4476-4481.
47. Valsami, G., Macheras, P., and Koupparis, M. (1991) Binding study of the fluorescence probe 1-anilino-8-naphthalensulfonate to human plasma and human and bovine serum albumin using potentiometric titration., Pharm Res 8, 888-892.
48. Takehara, K., Morinaga, Y., Nakashima, S., Matsuoka, S., Kamaya, H., and Ueda, I. (2006) Classification of the binding modes in bovine serum albumin using terminally substituted alkane analogues., Anal Sci 22, 1571-1575.
49. Gavin, P., M, P., DJ, F., and JJ, C. (2003) Minimizing Photobleaching of Blue Flourescent Protein (BFP) using Varian Cary Eclipse Fluorescence Spectrophotometer, In Fluorescence At Work: Varian, Department of Biochemistry and Molecular Biology Monash University.
50. Tsien, R. (1998) The green fluorescent protein., Annu Rev Biochem 67, 509-544.
51. Trott, O., and Olson, A. (2010) AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading., J Comput Chem 31, 455-461.
52. Whitchurch, C., Tolker-Nielsen, T., Ragas, P., and Mattick, J. (2002) Extracellular DNA required for bacterial biofilm formation., Science 295, 1487.
53. Nunn, D., Bergman, S., and Lory, S. (1990) Products of three accessory genes, pilB, pilC, and pilD, are required for biogenesis of Pseudomonas aeruginosa pili., J Bacteriol 172, 2911-2919.
54. Richards, J., Ballard, T., Huigens, R. r., and Melander, C. (2008) Synthesis and screening of an oroidin library against Pseudomonas aeruginosa biofilms., Chembiochem 9, 1267-1279.
55. Richards, J., Ballard, T., and Melander, C. (2008) Inhibition and dispersion of Pseudomonas aeruginosa biofilms with reverse amide 2-aminoimidazole oroidin analogues., Org Biomol Chem 6, 1356-1363.
56. Besenicar, M., Macek, P., Lakey, J., and Anderluh, G. (2006) Surface plasmon resonance in protein-membrane interactions., Chem Phys Lipids 141, 169-178.
57. Homola, J., SS, Y., and G, G. (1999) Surface Plasmon Reonance Sensors: Review, Sensors and Actuators B 54, 3-15.
58. Schönbrunn, E., Eschenburg, S., Krekel, F., Luger, K., and Amrhein, N. (2000) Role of the loop containing residue 115 in the induced-fit mechanism of the bacterial cell wall biosynthetic enzyme MurA., Biochemistry 39, 2164-2173.
59. Petrash, E., Fesenko, I., Prymark, I.,Ivashenko, R., Salamatov, V. (2009) Graphpad Prism for Mac OS X Version 5 Software Mackiev
60. Berry, J. Unpublished - The crystal structure of a domain of PilP from Pseudomonas aeruginosa
