Sofie Clais' PhD defense | The applicability of virulence inhibitors as a therapy for porphyromonas gingivalis infections
P. gingivalis. Etiological agent for periodontitis
Porphyromonas gingivalis, a Gram-negative anaerobe bacterium, is thought to be a major etiological agent associated with adult periodontitis.2Ingrid De Meester chairman: We're gonna start with professor Adyary Fallarero external jury, please.
Adyary Fallarero external jury ⤴So, hello, good afternoon, Sofie Clais defender. I should start by saying that I am really impressed with the presentation, although it was not a language I understand. But it was very well put together, and I think it's pretty much a reflection of your thesis altogether. I think your thesis is a really nice crafted document. And for a reader that is not tied or related to the topic, it's very easy to follow. So, one can only assume that, for a person, to be able to put together this document, with this attention to detail; I can only assume that this is the same way that you have performed your experiments, because, it's kind of generally a reflection of how a person acts, it's how a person can also write, so I just have that I'm really impressed. I think that you did a really good job in putting together this document. So the questions I have are more to know your opinion, to see how much you have grown as a student; how much you have been able to understand, what you have been studying. So, I might not know the answers to some of the questions, but I just want to hear your opinion; and make it more in a kind of a discussion manner, if you can say that. So, let's start with the title. When I started the thesis for the first time, I had only come to the rationale part3, where you were kind of questioning what you want to do within the thesis. Pretty much your main question was whether this DPP4-inhibitors could be applicable as virulence inhibitors against p. gingivalis. Then I kind of started to think: well, it could have also been possible to use the title "The applicability of DPP4-inhibitors as a therapy for p. gingivalis", to make it more specific. I'm not saying that I agree with that idea, I'm saying that this was my first reaction. But then I came to the realization that actually, this is a really good title. "Applicability of virulence inhibitors", so a bit more general then just the DPP4-inhibitors. Can you say why? Why do you think that this was a better title, than just making it specific for DPP4-inhibitors? Yes. The reason why I chose ... First it was "DPP4-inhibitors", but yes, I wanted a more general title, to cover the entire thesis. Because we also looked at the virulence capacities of DPP4, so it wasn't only the DPP4-inhibitors, but it was more than that. It was more than only just the inhibitors. And also, yes, I think that covers everything. We wanted to be broad, because the testing of the inhibitors was more a smaller part. We did the development of the models and the role of DPP4, and that's why we chose to take "virulence inhibitors" and not "DPP4-inhibitors"; although we only tested "DPP4-inhibitors", but ... Yea. But I ... you've said it, exactly. You also developed methods, you also developed a model for biofilm formation; so if someone is interested in the future of looking for antibiotic compounds, so then, they will be able to do that, based on the work that you have done. I kind of really saw that, when you started the thesis, you had like tiny problems at hand: you had to find new therapies, and you didn't have the method to find the therapies. So it's a huge task what you had, actually, on the taken. And I agree: this is a better title. It's "applicability of virulence inhibitors", and for those that will continue the work, then they have tools to look for virulence inhibitors, not only for DPP4-inhibitors; in the future. So I think it was a good title. Then you talk about p. gingivalis infections, and you have said there that it's affecting 90%4 of the Belgian population, as ~ periodontal disease. I think it's pretty much the same worldwide. What is the current treatment for p. gingivalis? If you have an infection, that it's ... Yes, if you have really an infection with p. gingivalis, it are ... first it's all the necessary depending if its gingivitis or periodontitis. One of the therapies is to disturb the biofilm, so then, the dentist will do that; using root scaling, it's called.
Periodontitis prevalence worldwide
A recent large-scale meta-analysis estimates that 11.2% of the global population is affected with periodontitis.5That's not pleasant. So; no, that's not very pleasant. Or surgical interventions, that's a therapy. But it's always in combination with treatment with antibiotics. And as antibiotics for p. gingivalis, they often use metronidazole or penicillins like amoxicillin; or azithromycin they also use; or quinol also. That depends a bit, but it's usually a combination: antibiotic with an additional therapy, and of course also: the good oral hygiene, because that is also essential in preventing those diseases. Yes, exactly. Good. So then, I don't know if it's ok, but let's go through different pages; and then I can ask specifically, in those pages. So, in the first chapter6, in the introduction chapter, in figure 1.27,yea, in page, at least in my version, it's page 23, yea. This is a really good figure that shows the 3 most important quorum sensing mechanisms in Gram-negative bacteria. And this is something that I really like, that you took up a lot in the introduction: you talked a lot about quorum sensing. But then, somehow, it wasn't that much taken up, in the rest of the thesis. You kind of mention it briefly, but not with so much of detail. So I just want to ask: which quorum sensing mechanisms is p. gingivalis using? So, for p. gingivalis, it's the second mechanism. So, with the LuxS and the autoinhibitor 2 ... autoinducer 2, that is used for p. gingivalis. But we didn't further focus on those quorum sensing mechanisms; because, yes, we focused on the protease mechanisms, but, yes. No, yea, that's good. And do you think it makes sense that they have this mechanism, and not the autoinducer 1 mechanism. That p. gingivalis uses that mechanism; and not autoinducer 1, which is kind of the typical for Gram-negative. Yes, I think so. First of all, it's a Gram-negative, so it's typically ... or it's also associated with that mechanism with p. gingivalis, with Gram-negatives. But, secondly, I don't know whether the other mechanism haven't yet been demonstrated for p. gingivalis; or he lacks it. But, I think, bacteria can have different quorum sensing mechanisms; and that's one, so: it was shown for p. gingivalis, that he had the second mechanism; so it's completely logical. Yes, you are completely ... Yea, that's right. I meant it more like: if you think it makes sense in the dental plaque, in the place where it lives. p. gingivalis is living there, associated with other microorganisms. Yes, of course. First of all, the quorum sensing is important in the biofilm formation. So, it lives as a biofilm; so, yes, therefor we see quorum sensing. On the other hand, quorum sensing is also more important in general for virulence production; so not only focused on the biofilms. But, then, thirdly; why we see AI-2 and LuxS for p. gingivalis, maybe the other bacteria in the oral biofilm also use that pathway frequently, as a quorum sensing mechanism. Yea, that's exactly right. I mean, there are so many different kinds of bacteria in oral plaque, and that's the kind of mechanism that can allow interspecies signaling. And you have, at some point, mentioned; when you were describing the way in which the oral plaque is formed, that Fusobacteria, for example, forms. It's a bridging organism and it's been shown that autoinducer 2 produced by Fusobacterium, then can be used by p. gingivalis. Yes. So, it's a very good convenient mechanism so that both Gram-positive and Gram-negative can communicate and can live in the oral environment. But then, the reason why I took up this interspecies signaling, I know that this hasn't been the topic of your thesis, but the reason why I took it up is, because one of the main discoveries, or one of the main findings of your thesis, is that DPP4 is up-regulated in biofilms, and that's clearly demonstrating here. Do you think that there might be any type of connection between quorum sensing and up-regulation of proteases or up-regulation of DPP4? Yes, I am quite convinced that quorum sensing pathways, that they up-regulate proteases like DPP4. So, there is also a study that has been done, which looked at up-regulation of enzymes, specific on p. gingivalis. And they saw different kinds ... some enzymes, or some genes were up-regulated, others were down-regulated. So, it seems quite logical, I think, that the quorum sensing may affect the DPP4-production. Yea, I think that's right. And it has been shown also that LuxS-mutants, they have lower production of many different proteases; including gingivitis, that you were talking about then in the thesis. So, I am just saying this, because this is another kind of direction that points out to the importance of your results. If quorum sensing is also linked to protease production and it's regulating that process, it must be associated that is important for biofilm formation. And it's kind of in the same line, that it's corroborating what your results have indicated, so that's the reason why I brought it up. Thank you. Ok, so then, we can go to ... We keep on that chapter6, and then, in page 27; you were mentioning some novel therapies, for example antimicrobial peptides. Antimicrobial peptides, is one of the, sort of hot topics nowadays, in antibiotic therapies, and in general: in antimicrobial therapies. And you mention here, that one of the disadvantages of the AMPs is that they are unstable to proteases; of course, that makes sense. Can you think of any other disadvantages of AMPs, that are relevant in the context of your thesis, I mean? Yes. First of all, they often get ionic. And that's already a problem, to get through a biofilm. So, it will be difficult to apply antimicrobial peptides for treatments of biofilms. Also, I think, the specificity is a bit too low, what's a big problem. And also, I think, that it costs a lot to produce them. So, I think, that are 3 disadvantages of antimicrobial peptides that make it difficult to find one, as a good therapy. > Yea, and they are not very effective against Gram-negative bacteria, so that's another thing why they might be good for many different infections, but not from Gram-negative. So, even though, it will be an ideal situation; they ... at the beginning, I think, many of them were very monogenic, because, of course they are protein. But nowadays they have ... it has been able to synthesize really small ones, so productions are also not an issue any more. But, still, they are ~ Gram-negative, it remains pretty low. So, that's why; it's kind of a ... they are important; but they won't solve the problem of p. gingivalis, in this case. Ok. So then, then I think you mention, when you were presenting, you sort of touched upon this topic, but ... I have diabetes mellitus, so every time that someone talks about diabetes, I have to talks about this. So, how relate it? Why is it that ... do you think p. gingivalis is a cause, or is it like a consequence of many of the issues that you have in diabetes. So, can you explain? Yes. There's a lot of discussion about it. So, first of all, it's a risk factor. So, often, when you see p. gingivalis, or when you see periodontitis, you often see, also, other diseases like diabetes. So, but, they showed the association. But the problem is: both diseases; or a lot of diseases, like cardiovascular diseases; are so complex and caused by different factors, that's its very difficult to find 1 cause or 1 relation. So, there has never been a causal relationship that has been demonstrated. So, what I can say is, maybe, if you have periodontitis with p. gingivalis, there are studies that demonstrate, so: the invasiveness of p. gingivalis. It can go to the liver, and there it can interact with the process of the glucose mechanism. So, the invasiveness, first of all: p. gingivalis makes that it may ... yes, I would say: induce the diabetes. But, on the other hand; you see, for diabetes patients, the problem is there: they may have a loss in bone density. And when you see to the periodontitis process, or the disease of periodontitis; it's an affection of your bone ... the bone tissue. So maybe there is a correlation, in that way that it provokes ... or that the causal relationship, I will say, that it induces the periodontitis. So, it's not ~known~. And then, as a third thing, I wanted to say, maybe it's not both that one causes the other. But maybe it's more big environmental factors, that make that both diseases can be associated. So, when a person doesn't pay attention to good oral hygiene, and doesn't eat healthy; and eats a lot of sugar; yes, than, of course, you have automatically more chance of developing diabetes, and also: it increases a bad oral hygiene, or it increases the dental plaque, and induces periodontitis. So, there are 3 mechanisms, or hypotheses, I think, that can be followed. Yea, that's very convincing. I don't know if you know my questions in advance? Because, you are not looking very surprised when I am asking you. So, it's kind of ~. Yea, she is very well prepared. So, then, we are coming to chapter 38. And then it's the first time when you're trying quantify ... the first problem that you have that you were comparing different kinds of quantification methods. And why; if we look at figure 3.59, this regression analysis that you have made for the 3 methods; and why did you choose this concentration range from 106 to 109? You have kind of ... why did you choose this one? Yes, so the reason why we focused on more highly concentrated suspensions is because we saw those repeatability problems during we measured the DPP4 activities for highly concentrated suspensions. So, for the fractionation protocol, it was necessary that we had enough bacteria; but that suspension need to be quantified; so therefor we looked at highly concentrated suspensions ranging from 106 to 1010, to be sure that we covered the whole area, and to find the best quantification method for our specific problem, actually, in the lab; of quantifying highly concentrated suspensions. Ok. Yea, that makes sense. So, you have said; I think it was written on the text; that qPCR has a detection limit ... was it 103, could be? Yes. Yes, also. Yea? So, what about ... and turbidity has a very bad detection limit, you can see it here ... what about viable plate counts? What's the detection limit of the viable plate counts? Yes. The viable plate count, also. Actually, the detection limit is in theory 103, because when you look at the process of diluting, you can't go lower than 103. But the problem with p. gingivalis was, due to its difficult ... Yea, they aggregate. ... to culture, also aggregation; and due to the difficult culture conditions; it was very difficult to measure the lowest suspensions. Ok. That makes sense, ok. So, then we go to chapter 410. You have said here, in page 96, or at least in my version it's page 96; that for anaerobic metabolism, it can be that you required more specific dyes, to measure; for example if you are looking at viable cells. Yea? Uhum. Yea. So like, that you should use a ... specific redox dyes. So, you ... when I interpret from here, is that you mean that you cannot use this sort of more typical redox dye like MTT or resazurin. So you don't think that we resazurin to measure p. gingivalis? No, we tried it. So, it's not applicable for anaerobic organisms. So, because it's also already a type of ... in a different state when it's in the anaerobic environment. So we tried it, but it didn't work. The same for the XTT and the MTT. We hadn't good results with that. No, because it was just really interesting, because we have used resazurin too, with aggregatibacter. But, and I have seen in the literature that some people ... it shouldn't, in principle, be possible; but I have seen papers in which they have used this. So I was just interested as to why you have said. But maybe it is that people grow the bacteria anaerobically, and then they do the experiment aerobically. Yes. I don't know, but ... I think it differs with the bacterium, first of all, maybe; but also the conditions in which you culture it; maybe. Sometimes it's not really an anaerobic environment, I think. Yea, it can be ... or it's micro-anaerobic ... Yea, it's microaerophilic, so ... Yea, microaerophilic conditions. So, it can depend. True. Ok. So, then we go to chapter 410, that's my favorite chapter. And, then there is this figure 4.311, in which you have ... it's when you started optimize the conditions for biofilm formation in microtiter well plates. And you have shown that the bacteria grow pretty nicely, or it forms biofilms up to 72 hours. And then it kind of reaches a plateau ... Yes. It sort of not ... it doesn't continue to grow. So, this was done in microtiter well plates, right? Yes. So, what's the surface in microtiter well plates? The surface. I think that are polystyrene plates, in the well plates? Yea. And what kind of surface is that? Is it, like; positively charged, negatively charged? Is it ... I think positively charged? It's more hydrophobic, but ... Ok. So, it sort of binds. It's slightly hydrophobic, so p. gingivalis binds very nicely then to polystyrene, I understand. Do you know if it also binds easily to other kinds of artificial surfaces that you can use? Yes, I think it can. It'd be different depending on which kind of plate you use. So, we didn't measure or evaluate other types of well plates. But I agree that it can be different. So when would have changed a plate to another type of well plate, we would have investigated it, but we didn't. We always used same plates. But it's true that it can have an effect. Also, when you would cultivate the p. gingivalis on glass plates, it will be a different ... in biofilm formation. Did you try in glass, just out of curiosity? No, we didn't try it, no. Ok, no. Ok, because that's a very polar surface, so it might be different. Yes. I actually ... I'm not ... my comment was, because I was really surprised, when I started to check, that some other people have tried to grow p. gingivalis, for example in hydroxyapatite or in titanium disk. Yes, yes. And it seems that the kinetics is exactly the same. They get to grow very nicely up to 72 hours, and then it drops down or then it gives a plateau. So, this is pretty much the same as what other people have found. Which makes us think that, then, this is a very good model. If it's similar to the attachment that you can see in, for example, in implant surfaces that you use on the bone. So, it's another thing that points out to the validity of the model that you have established. Ok. Then, in that very same chapter10, I'm page 10412; yea, I don't know if, in your case, it's the same. You have said here, that when you were trying to use the DiMethylMethyleneBlue as a stain for the matrix. You have said here that "the composition of the p. gingivalis biofilm matrix is not yet fully understood" and pretty much this probe interacts with polysaccharides, with these sulphated polysaccharides. Which then could make us think that the matrix doesn't contain these sulphated polysaccharides. Can you think of an experiment in which you could figure out what kind of matrix you have in p. gingivalis? A simple experiment. Yes. For a simple experiment, I think you could stain it. For example, ~if you want~ ... But it's difficult, because you will always also stain the bacteria. And you want to specifically know what's your matrix, so I think you have to make the difference. You have first to stain a biofilm with a stain that specifically stains protein, or saccharides. And then you have to compare it with a planktonic variant, to be sure of what is composed your bacterium and of what is composed your matrix. So, yes, you can use staining methods. Maybe labeled anti-bodies, I think. Or can you do a more destructive manner, and you can determine the matrix using other techniques, like mass spectrometry; I don't know, but it's certainly worth investigating. Yea, I was thinking on something even simpler than that. Can you think on: how would you know if the matrix is composed mainly of protein, for example? Because there are biofilms that produces matrices that are mostly composed of proteins. So then polysaccharides are not the essential element there, or then they have maybe more of the ~extra-cellular~ DNA. Can you think on a simple way in which you could figure that out? I think at stains like Coomassie Brilliant Blue, or ... Yes, that specifically interact ... or kit to detect protein, so I think that's ... Yea, you can; one way, for example; is that you can treat the biofilms with proteinase K, and see if they resist that. Yea, or also that. Yes, then you have it directly. Or then DeoxyriboNuclease 1, and then see if they resist that, for example. That's true. Yes. So, that's just a question, just to ... Ok, so then, there is this table. I think it's my favorite table in the thesis. It's table 4.513, in which you are validating the optimised crystal violet protocol. You have pretty much used the same kind of things that we do in my lab. We just call in it a different way. For example, when you refer to this "intraday variance", we call that "plate to plate variability". And then, when you measure the intra-plate variability, then we call that "well to well variability", but it's pretty much the same. We have, for 5 years, we always measured every time that we were optimizing assays, we were always measuring: "well to well variability", "plate to plate variability" and "day to day variability". And last year, we had an American professor visiting our lab, that has been involved in many, many different kinds of large-scale optimization of assays. And then, she simply asked, like: "Ok, that's nice. You have measured well to well, plate to plate and day to day. What about operators variability, persons to persons variability? Have you ever measured that?" I didn't do it for p. gingivalis. It's true, it's an important factor that determines your variability. Because, the way that you ... the rinsing method, for example, using the vacuum pump. It's true that you have to be an experienced biofilm researcher, I would say ... Yea. ... to do it properly. Because you can easily absorb the biofilms with the pump. So, I didn't do ... we didn't try it with other researcher. But, I'm quite sure, if someone else would do it, an experienced one; maybe we would have the same results. Yea, but students do it the first time ... It can be very difficult. ... it's completely different, and ... Yea, we have never measured that. Then, what we started to do, was like: we started to follow how the protocol will go, as the person will start getting trained. And then it got really better. It was like ... we do it by measuring the parameter ... I will ask about it in a minute. But, that's a very interesting thing. We didn't do it like that before, but the thing is that regulatory authorities; when a method is put up for acceptance, in the revelatory's policy; that is one thing they always ask. If there are differences between operators. So, just to think about this, for the future. Then, what about the detection limit of crystal violet. Did you measure what is the detection limit of crystal violet? How much biofilm the crystal violet can detect? We didn't looked at how much biofilm. So, I think we already saw the distinction between the weak biofilm formers and the moderate / other. Yea, that's good. So, that's already a good point, that we took different types. But, I cannot surely say how much, or how low it can go. What we did was a linearity curve, for the crystal violet itself. But that hasn't much to do with the biofilm amount, I think, so ... It's quite laborious, because you have to scrape the biofilm all the time, so it's not ... Yes, I think so, it's not easy to find it out, I think. What about the thickness of your biofilm? Could you see, could you calculate what the thickness of the biofilm was? ~ because it's quite all ~, in that sense. Yes, I didn't calculate how thick he was. I can say that you could see it quite well, with the different types. So you really see only a bit of turbidity, I would say, on the surface. Or for the P. gingivalis W50's tracer, it was also; it was more dense already. And the 80cc was really a thick biofilm. So, I saw it visually. But we never measured it how thick he was. That's good. So, when you have mentioned at the end, it's one of the important things of your work; it's that you can ... all of these methods that you have developed; at least these 96-well plates methods; can be used as a high-throughput screening method. And if you intend to use something as a high-throughput screening method, there is a parameter that always screeners measure; and it's scalled screening window coefficient. And the screening window coefficient, pretty much, what it does; is that it takes into account the differences between the mean of the maximal signal and the mean of the minimal signal, and then the dispersion of the data. So, just to have an idea of what the signal window of the assays is. So, that was why I kind of ask you to put the background values. Can you tell me, more or less, around what observance value, were like your background values; generally speaking? My background values are usually 0.08. Yes. I made a guess, that it would be around 0.1, which is even a little bit higher than that. So, assuming that if you would have a minimal signal of 0.1, and the standard deviation would be like roughly 10%; well, let's assume that. The set-up prime of your essay ... with this screening window coefficient of your assay; taking the worst of your plates ... Yes. ... it's over 0.45. And screening window coefficients, typically speaking, it's very good when you have over 0.5, but for cell-based assays, it's very difficult to have 0.5. So, it's very good if you have more than 0.4. So, it's a very good assay; so, I just wanted to add that. If you would have calculated this, then it would be another element that would justify that this is a very good assay. And for all this complexities of growing anaerobic bacteria, I would say that it's an excellent essay to use for high-throughput screening.
I presume Adyary Fallarero external jury talks about high-content screening, regarding the acceptability of a Z-factor ≤ 0.5
While Z' > 0.5 has become a de facto cutoff for most HTS assays, 0 ⟨ Z' ≤ 0.5 is often acceptable for complex HCS phenotype assays, since those hits may be more subtle but still valuable.14Ok. And then, the last question I had about this, is that: did you measure the tolerance of your biofilms? Because you used then metronidazole, I think ... Yes, yes. ... in the last set. Did you measure how much higher concentration of metronidazoles are needed to kill those biofilms, as compared to planktonic values. Like, something, like, comparing InhibitorConcentration50's. Yes; no. We didn't it for the metronidazole. Or for any other antibiotic? No. Not for the antibiotics. So, when we would have treated them as a control, you would take high concentrations enough and on a long time. So, what I showed you, was an example for the LIVE/DEAD® (BacLightTM) kit, that we used ... And what was the concentration there of the metronidazole? By heart, I don't know if it's measured in the methodology section. No, it doesn't matter, it's ... 32 micromoles, but I'm no sure, I have to look it, but it will be ~. Ok, yea. But that's the high concentration anyway. Ok, good. So, we are getting maybe to the end. Then, in the in vivo assays, in chapter 515; you had a table16, a really nice table, in which you have summarized all the in vivo oral infection models of periodontitis. And there you have the mice model, which I assume, this is the model that ... is not the model that you have used. No, it are only oral infection models. Yes. Yea. Yes, can you explain how these methods go? Like how the models ... how this model is what ... how do you induce the infection? Yes. So, in general, when you have that kind of animal models with oral periodontitis; you use ligatures to induce. So, for one thing: you can only put the ligature. That accumulates the bacteria around the ligature, and induces the periodontitis. Or you can, before you can treat, you can, yes, put the ligature in your bacterial suspension, or the bacteria that you want to investigate; and then you put it around the teeth. And that will induce more periodontitis, when it's covered a bit in the animal model; it'd go faster then when you would wait naturally. But another way of inducing the periodontitis, or getting the infection; is by using viscous suspensions, so with viscous bacterial suspensions. So then you have to put, for example, 2 times a day, the bacterial viscosity ... Directly on the teeth, or? ... directly on the teeth, yes. In order to create there infection or colonization with that type of bacteria. So, if you were to put exactly, because you ... I like this table, because you have the positive; and the pros and cons of every model. If you were to do the same for your murine back abscess model, what do you think is positive and negative of that model? The back abscess model, yes, it's different; it's not in the teeth, so first of all, you don't have the real situation, I agree with that. But, they used it, for example, to see whether DPP4-deficient strains or ... induce virulence. So, it's a good model for a first screening of virulence, but it's not the most optimal model to look at the real situation of periodontitis. And you also don't take into account that there are other bacteria; so only focus on 1, but that can also be an advantage, that you only have 1 pathogen, and that you don't have the problems with all the other bacteria in the mouth. Yea, the complexity of all the others. Ok. So, then we go to chapter 617, and here, I think, one of the most really elegant experiments that you did was, when you compared the DPP4 activity in both planktonic and biofilm phase. And, so, I want you to explain to me how you did that experiment, just to reflect with you on that experiment. Like, how did you do it? Did you grow ... explain me how exactly you did that experiment. Yes. So, we took a 24-well plates, and we grow there the p. gingivalis biofilms. And to compare, because the circumstances had to be the same for biofilm bacteria and planktonic bacteria; we took the supernatant, that we used as the planktonic ones. And then we scraped only, we removed the supernatant, and we scraped the biofilms, and we measured in the biofilm and in the planktonic one; we fractionate it through the protocol; and there we measured the DPP4 activity. And to quantify them, we used both PCR and the viable plate count method to quantify them. And then we compared. So we saw, for the both quantification methods, when you calculate the DPP4 activity on the number of bacteria, we saw an increase in DPP4 activity for the biofilm ones. Ok. So, these bacteria that you have on the planktonic phase there, when you start to seperate both phases. That planktonic phase, where do you think these bacteria come from? Do you think these are bacteria that never attached, or are it those bacteria that got detached from the biofilms? It can be that detached. But, I meanly think that they never growed as a biofilm. So, even when they detached from the biofilm, I think the amount of planktonic bacteria is high enough, to call it a supernatant, or to call it planktonic bacteria. So, I think that's a good comparison. On the other hand, we also did it; we compared ... it's not mentioned in the thesis, but we also wrote it once ... only planktonic, without shaking. And there also we saw lower activities. But, to have the same conditions; because, maybe, your shaking condition can also have an influence on the virulence, or on the production of the protease; we took the supernatant and not planktonic cells that were grown on the static platform. This is exactly how we do these kinds of experiments, so I'm not questioning anything about how this was done. But, we always wonder; and this was something that, for many years, it bothered me; that I was almost sure there was always detaching cells from the biofilm. Because it's part of the normal life. And you have done it in 24-well plates, was it 24-well plates? So there, there is much larger area. Yea. In 6-well plates also, but ... Yea? Yes. So, they have a larger area; but when you do it in 96-well plates, as we do it, the area is so small that at some point, when the biofilms are formed, then they start to detach. So you can't, I think you can never get rid of these detached cells, they are always there. And unfortunately, last July, there was a paper18 published that ... I was like: why didn't we do it? But ... there was a paper published showing that, actually, detached cells or dispersed cells; which we always assumed that were immediately going into the planktonic phase and they were like planktonic cells; they are having a specific state, like a very specific functional state that differs from planktonic cells and differs from biofilm cells. So, what we learn from this paper is that, I think, that it's; we have to be really kind of careful in the future, when we do these kinds of experiments, just because there will always be some small population of dispersed cells in these kinds of experiments. And as I said, this is how we're doing the experiments, but it's just that maybe we are making assumptions that the whole oral population is in the same functional state. And it's not, because you have a dispersed population and then a planktonic population. So, that's just one small note that I want to make. Like, I think, I agree this is the best way to do it. And you can't ... if you do it with seperate populations, then they are not comparable. And you are doing it exactly from the same population. So, it's always a trade-off; of to what to do, in that scenario. But it's just how it goes. And then, yea; I think that we come to the end. Then, let's say that you are becoming a supervisor now; that you are finalizing your thesis, you are a post-doc. And then you get a student, and the student comes to your lab; and then you have to guide that student and tell him ... tell this student what the student will do during his doctoral studies or her doctoral studies. So, what do you think that this will be a future direction for your thesis; like, what can you say? First, I can tell you that this is what has been done, and this is what needs to be done; from this point forward. Yes, I would investigate a lot of new things, I think. First of all, I want to further study the effect of the DPP4 on different cell types. So, I think that's interesting to find out; whether there is an effect of p. gingivalis DPP4 on the immune process of the host. But, on the other hand; I think it's worth to test new inhibitors, or alternative inhibitors, more inhibitors in the biofilm model that we have and in the in vivo model that we have. But also, when there is time, when there is money; when it all went fine; I would also look for, maybe, an oral periodontitis model or the multispecies biofilm model. So, maybe in the first option, the multispecies biofilm model, because it's more easy to implement than the oral periodontitis model. Thank you. It has been really, really good. Ingrid De Meester chairman: Thank you for the nice discussion. And perhaps in sake of time, I will just ask 1 question of professor Wim Teughels external jury, absent, and then we continue with the other members of the jury, who are present. Ok.
Wim Teughels external jury, absent ⤴ read by Ingrid De Meester, chairmanProfessor Wim Teughels asked us, to just congratulate you. He enjoyed really reading your thesis, and he sais it reflects a lot of work, very careful work. But, and now he has a few questions, and 1 of them continues on the discussion we just had. And he writes: "Well, it's currently clear that oral diseases, and specifically periodontal diseases. And so, how does your protocol of just taking 1 target of 1 of the microbes, how does it fit in with this multi-microbial view that one has today?" Okido. Show me the comment on that. Yes. So, I agree with him that we should have ... that it will be worth to use a multi-species biofilm model to, for example, test the DPP4-inhibitors. Because, maybe, I think, that the DPP4 can be involved in multi-species interactions; in different ways. So, as I told, maybe DPP4 has a role in the adhesion process, maybe it allows ... It has been shown that DPP4 can adhere to the matrix component in the mouth. So maybe the DPP4 at the outside of the bacteria, can adhere to other components of other bacteria in the multi-species biofilm. On the other hand maybe the DPP4 plays a role or provides also nutrients for other bacteria that live within the multi-species biofilm. So, it's worth to take a multi-species biofilm model, to investigate DPP4 inhibitors and to investigate the role of DPP4, I think. And maybe then you can use DPP4 knockouts, for example, and combine them with other bacteria, like aggregatibacter for example, and you can cultivate multi-species biofilms with DPP4-deficient and DPP4-positive strains; and to see whether there is an influence on biofilm formation. I think that are all possibilities that are worth to investigate in the future. Ok. The second question was about the abscess model you used, but you answered already on 1 of the questions of the previous jury member. Yes. So, this is, I think you just answered this one: "The advantages and drawbacks of your model, just compared with other models; which are more, let's say, interesting for dentistry." Ingrid De Meester chairman: So then, I think I can hand over to colleague Pieter Van der Veken internal jury. Pieter Van der Veken, please.
Pieter Van der Veken internal jury ⤴Thanks. So, although I speak Dutch, Sofie Clais defender, I will continue in English; just as the other jury members. First I would like to congratulate you, because ... and I have already done that, but I'm going to do that again today ... it was a very clearly written thesis that you have submitted, and it very well reflected what you have been doing. And I really enjoyed reading it. Thank you. So, that said, I will continue my questions now. And my first question is about 1 of the things that you already have been showing on the screen: the prophylactic use of virulence inhibitors. This is, at the same time revolutionary I think, and a bit strange. I will specifically refer now to what you have written in your thesis: that you might expect less resistance problems with virulence inhibitors, than with classical antibiotics. And that's certainly important in the framework of prophylactic use of virulence inhibitors. Could you explain me why you expect less resistance problems with virulence inhibitors, and whether that has been proven in reality? Yes. So, there is also discussion about it. First of all, I think, personally, that resistance will always develop for every type of antibiotic or virulence inhibitors. I think there will be no magical bullet; I hope so that once researchers find it; but I think it's very difficult to find one that doesn't induce resistance; because resistance is simply the way that bacteria adapt to unfavorable environments. But there are so many kinds of virulence inhibitors, and I also think that the chance of developing resistance differs among these virulence inhibitors. So, first of all a virulence inhibitor; because it causes less selective pressure on the bacteria, the resistance development would decrease or slow down compared to the normal antibiotics. So I believe that theory, that it's possible. On the other hand, they're also ... depending on which inhibitor you choose ... they call it environmental specificity. So, that's a way why the resistance would less develop. So, for example, in the thesis it's mentioned19: there is staphylococcus aureus, it's present as a commencile in our body; but it can also induce disease. So, it can produce the enzyme crtM, that produces staphyloxanthin. So that protects the bacteria from an attack, by reacting oxygen species by neutrophils. So when you add a virulence inhibitor, at the site of infection the staphyloxanthin is not produced anymore; so the ROS will kill the bacteria. But at your commencile site, there is no attack by the neutrophils, so there you don't develop the resistance, because the bacteria don't need it. So, that's a way of decreasing resistance development. I know that it's a bit ... there is a lot of discussion about it ... but I think it's possible to find 1 which decreases or slows down resistance development. Ok, yea. You could go on discussing that for hours, I think. And that's not what we ought to do. A second question would be related to the clinical isolates K4 and K520 that you have been using. Apparently these were the most virulent isolates that you have been evaluating; and you were able to correlate, for these at least, virulence with the amount of DPP4 that was present. Now my question is: have you been looking at other proteases that are being secreted by these bacteria? No. So, we only focused on DPP4, yes. Ok. And do you know how you would do the experiment to find out whether there are other proteases in active, or in present? Yes. If you want to test other proteases, you can use other substrates, like we did for DPP4. But you can also do reversed transcriptase PCR, so then you're going to look at the expression on messenger-RNA level of different proteases; I think that's a good and fast method to evaluate, or to screen, actually, the bacteria for a lot of proteases at the same time. I think it sounds like a very good idea to try that, because I would be surprised if only DPP4 would be up-regulated in these virulent species. If you look at the genome of p. gingivalis, how many proteases are there? I don't know. A lot, I think. A lot, yea. I have no exact numbers of how many proteases there are. Maybe, some are not known yet, so. Maybe it's worth to investigate which proteases they may have. Ok. And if you look at the families of proteases that are present? Like, for example, you have serine proteases, cysteine proteases. Which ones are present in the genome, or you haven't really ... 1 of the proteases that is present are the gingipains, I think that's also an important one; that are arginine-specific gingipains, that are included. So, also a carboxy metalloprotease. So I think there are different families that kind be found for all kinds of bacteria; but also for the p. gingivalis. Because it's an asaccharolytic bacterium, I think he would have a lot of types of proteases to fulfill the nutritional requirements. Yea, that sounds reasonable. That's right. Have you been checking whether p. gingivalis is capable of breaking down collagen on its own? We wanted to optimize a collagen degradation model. So we started with it, but it's difficult; because, first of all, we start to combine the collagen with the cell; so we chose a medium. First of all, it should be mentioned, when you combine it's ... yea, we focused on combination with the cells ... the cells grow aerobically, but the bacteria anaerobically. We did a lot of studies, but the problem with the collagen degradation is, when you look at DPP4; it cleaves dipeptides, so the break-down products are ~the peptides~, and that's difficult to measure. So, when you would stain the collagen gel with a sircolTM assay, you stain ... you will miss the effect of DPP4, I think; and you only look at big fragments, while DPP4 degrades small peptides. So to optimize that, you should combine it with mass spectrometry, for example, to specifically measure the dipeptides of the collagen degradation. Ok, yea. Next question ... and I'm not going to ask too many questions ... is: if you look at the assay in which you quantify the mass of the biofilm, you use crystal violet. And do you know the interaction on which this assay is based, between the dye and ...? Yes, the crystal violet interacts with both the bacteria and the matrix, with negatively charged components of the whole biofilm actually. And if you look at biofilm mass chemically, what negatively charged components are there? In the whole biofilm? I think a lot, there can be polysaccharides that are negatively charged; maybe depending on the pH of the environment, you can also have protein that are negatively charged I think. Or other acidic components. Or, in the capsule, there are a lot of ... D-mannuronic acid, for example; I think it's composed of. So that can also, I think, be negatively charged. Yea. And if you incubate with acetic acid, what is that for? So, the acetic acid in the biofilm staining protocol; that's for the crystal violet stains, and then it's dissolved again, yes, in the acetic acid. So it loosens the crystal violet from the biofilm. It's released, the bound dye is released. Yea. If you look at the biofilm inhibiting properties of compounds, you have evaluated diprotin. You have evaluated "diprotin"? Diprotin, the compound. Yes, yes. And you have also been using metronidazole as a kind of positive control. But, metronidazole is actually an antibiotic. It doesn't really affect biofilm mass, it just kills them ~. No, it was as a positive control to compare that we added both the diprotin A and metronidazole. So, we want to be sure, we want to see effect. So we add it to a growing biofilm, and we wanted to know whether the diprotin A has an effect on a growing biofilm. And to have a positive control we added the metronidazole as an antibiotic. So, of course; it's an antibiotic, so the bacteria are killed and that's why there is an effect of that. Are there any compounds know that specifically affect biofilm formation, but not viability of the cells? Yes, if you have a good quorum sensing inhibitor, I think it would affect biofilm formation without affecting the viability. But then you should look it up in literature: which compound specific that can target biofilm formation without killing the bacteria. Ok. Maybe a final question then. I had a bit more, but it's getting late. You have been discussing; that was also in my colleague's question here; the invasiveness of p. gingivalis Probably there is also other species that invade the body? Yes. Also from the oral cavity. Which other species are known to invade? I think ... other species that invade ... I think prevotella intermedia, also in dental plaque; can also invade. And the aggregatibacter actinomycetemcomitans, I also think he's important to invade the tissues. But there will be a lot of other species, I think, that can do that. And how do they evade the immune system? Evasion of the immune system, for example, by capsule formation. I think that's a mean important factor to evade the immune system. On the other hand, they may produce ... Yea, I'm not sure at the moment. They have other mechanisms to evade the immune system, but the capsule formation is an important one to circumvent that. But, for example; also you can interact with the reactive oxygen species, if there is an attack, intra-cellular; that it cannot degrade the bacterium, when he has a virulence factor that interacts with the ROS. Ok. Thanks.
Anne-Marie Lambeir internal jury ⤴Sofie Clais defender, I'm going to join my colleagues here on the front row by congratulating you: very nicely written thesis; very educational also, has educational aspects in it; and good presentation. And you may think, or the audience may think that you had to answer a lot of questions right now, the last hour or so. But we've asked a lot more questions over the years ... Yes. ... trying to get this thesis where it is now. So I do congratulate you. Thank you. Many of my questions, of course, have already been asked and answered. Answered very well, in some cases. So I have 1 question for you to finish off. And it relates to what you said in 1 of your very last slides: that you are ... that you feel there is a need to test more inhibitors. Ok, I will ask you now: what would be your ideal inhibitor? What would be the criteria that you would use choose an inhibitor or a set of inhibitors, to test in your system? Yes. So, first of all, very important, I think: a selective inhibitor that selectively acts on p. gingivalis' DPP4. I didn't mention it in the presentation, DPP4's are also produced by the human. So I think that's important: that you really specifically target p. gingivalis' DPP4. And when you then have a specific inhibitor, when you have it in sufficient amounts; I think it will be good to test him in high concentrations on the biofilm formation, in the abscess model. Also, it would be interesting to see the stability in the biofilm model, because the inhibitor is during 96 hours in the biofilm plate, which is not very ideal. So it would be good to look at its stability in that type of conditions. On the other hand, you can also test the inhibitor on the multi-species biofilms, if you would have 1. And it would also be good to evaluate; when you put the inhibitor an in vivo model, in a mice model; to see whether it can ... when you give it orally to the mice, it would be interesting to see that it effectively inhibits the DPP4 in the bacteria, yea, at the infection site. So that's also a point to consider, that you are not sure whether it targets DPP4 at the site of infection. So, a lot of possibilities, yes: multi-species biofilms, the in vivo model, mono-species biofilm, stability of the inhibitors. So, but you need a good compound at a high amount, so that you can do a lot of tests. ~ Thank you, Sofie Clais defender. Ingrid De Meester chairman: So Sofie Clais defender, your promotors Paul Cos promotor & Louis Maes promotor, they have told me that they have asked all the questions they had before, so that they don't have any questions left. I would love to discuss with you hours, but I just will finish off here. Also, by congratulating you. It was a nice thesis, and you answered, I think, very well on a number of questions.
1. Sofie Clais attributed the following sentence to Louis Pasteur: The role of the infinitely small in nature is infinitely great. As have many others; however, I have not found a reference to justify this. Its origin seems to be in the writings of one of his biographers René Dubos: The infinitely great power of the infinitely small. As can be found in:
▸ https://archive.org/..., René Dubos, Louis Pasteur. Free lance of science, Little, brown and company, p. 45, 1950.
Which itself seems to be a loose translation of Louis Pasteur's original quote: Du rôle physiologique, immense selon moi, des infiniment petits dans l’économie générale de la nature. As can be found in:
▸ http://gallica.bnf.f..., Louis Pasteur, Note remise au ministère de l'instruction publique et des cultes. Sur sa demande, in: Oeuvres de Pasteur, Masson et Compagnie, 1939, p.7, first: Bulletin de l'Institut Pasteur 21 (14), 31/07/1923, p. 537.
Which can be translated from French as: Of the, in my view, immense role of infinitely small bodies in the general economy of nature.
- 2. ▸ http://iai.asm.org/c..., Hisao Yagishita e.a., Infection and Immunity 69 (11), p. 7159-7161, 2001.
- 3. ▸ Chapter 2: Rationale and research objectives.
4. Sofie Clais defender said, in her Dutch public PhD defense introduction: There are no exact numbers about it, but it is estimated that 90% of the Belgian population has some kind of form of gingivitis. To which Sofie Clais defender added: And it is even estimated that 40% of the Belgian population has periodontitis. I presume Sofie Clais defender might be pointing to data discussed in the following study:
▸ P. Bercy e.a., Parodontale gezondheid en parodontale behandelingsnood binnen een steekproef van de Belgische bevolking, Revue belge de médecine dentaire / Belgisch tijdschrift voor tandheelkunde 3, 2002.
Since the following description of a study mentions the source above in its footnote: The team of professor P. Bercy has in 1997 conducted the first Belgian epidemological study on 402 staffs of the universitary hospital U.C. Louvain. They used the CPITN index and found periodontal pathology going from chronical gingivitis (99,9%) to deep pockets (41,4%).▸ http://vbtda3.axoclu..., Eric Thevissen, Parodontale screening met behulp van de DPSIndex, Contactpunt (march), p. 10-16, 2005.
This last source, however, mentions somewhat different number regarding the Belgian situation as well. Namely, numbers from a pilot project that ran from 1/2004-6/2004 in Limburg, which used the Dutch Periodontal Screening Index (DPSI). According to Eric Thevissen, the DPSI is a measure proposed after criticism of the CPITN index (which, according to the proponents of the DPSI, doesn't focus enough on gum recessions), posted in:
▸ ~, De CPITN Index, Het Nederlands Tandartsenblad 53, 1/1998.
According to Eric Thevissen, the DPSI was introduced since (translated from Dutch): It is the "total" loss of attachment, namely: pocket depth plus recession, which is determining for the remaining bone reserves around a tooth. The pilot study involved the screening of 819 patients during check-up in the dental context by 65 dentists, and obtained the following results, measured for 4 age categories:
- Has gingivitis: 24% (5 year olds) & 56% (15 year olds).
- ≥ 5mm loss of attachment: 37% (35 year olds) & 78% (55 year olds).
Another more recent Belgian study, which also used DPSI, concluded the following (details obtained via correspondence with author Jacques Vanobbergen : involving 3123 screenings between 25/1/2010-25/6/2013, where individuals might be screened 1-4×): Has gingivitis: 59% (12-16 year olds following Technical Secondary Education (Dutch: TSO) & Vocational Secondary Education (Dutch: BSO) in Flanders). ▸ http://www.glimlache..., Jacques Vanobbergen, Souriez / Glimlachen | Scientific rapport, ~, 1/6/2014, p. 107.
Other epidemological studies, regarding gingival condition in Belgian children, can be found in:
▸http://onlinelibrary..., Jacques Vanobbergen, Caries prevalence in Belgian children | A review, International Journal of Paediatric Dentistry 11 (3), 9/2011, p.164-170.
Within the framework of the Belgian interuniversitary "Oral Health Data Registration System (OHDRES)", in 2008-2010, there has been a clinical study performed by dentists, regarding periodontal health, by means of noting the highest DPSI-score per individual for a total of 1,563 observed participants, who are 15 years & older; which obtained the following results (please note that the following is no literal quote, instead, I have summed up the percentages of DPSI-indices 1 & 2 for gingivitis & the DPSI scores of 3-, 3+ & 4 for periodontitis):
- Has gingivitis: 48,62%.
- Has periodontitis: 18,38%.
- Has gingivitis: 52,7%.
- Has periodontitis: 17,9%.
▸ http://onlinelibrary..., David Spratt, Chapter 4.1. Dental Plaque and Bacterial Colonization, in: ▸ http://eu.wiley.com/..., Medical Biofilms | Detection | Prevention and Control, Wiley, p. 174-198, 2/2003.
- 5. ▸ http://jdr.sagepub.c..., Nicholas J. Kassebaum e.a., Global burden of severe periodontitis in 1990-2010 | a systematic review and meta-regression, Journal of dental research 93 (11), p. 1045-1053, 2014.
- 6. a. b. ▸ Chapter 1: Introduction.
7. Figure 1.2: 3 quorum sensing systems in Gram-negative bacteria.
- A: AI-1-based strategy using AHL molecules.
- B: AI-2-based strategy using QS signals derived from the precursor DPD.
- C: Al-3-based strategy using an autoinducer of unknown structure. Inner membrane, outer membrane.
▸ Chapter 1: Introduction.
▸ http://onlinelibrary..., Christopher Parker & Vanessa Sperandio, Cell-to-cell signalling during pathogenesis, Cellular microbiology 11 (3), p. 363-369, 2009.
8. ▸ Chapter 3: In vitro growth and quantification of P. gingivalis.
Parts of this chapter have been published in:
▸ http://onlinelibrary..., Sofie Clais e.a., Comparison of viable plate count, turbidity measurement and real-time PCR for quantification of Porphyromonas gingivalis., Letters in applied microbiology 60 (1), p. 79-84, 1/2015.
9. Figure 3.5: Regression analysis for the 3 methods.
- VPC versus expected values: y = (1.1380 ± 0.0609) x - (1.2340 ± 0.5194), R2 = 0.9645.
- Turbidity versus expected values: y = (1.3770 ± 0.2666) x - 4.346 ± 2.519), R2 = 0.7920.
- qPCR versus expected values: y = (0.8380 ± 0.0598) x + (1.5910 ± 0.5104), R2 = 0.9548.
▸ Chapter 3: In vitro growth and quantification of P. gingivalis
- 10. a. b. c. ▸ Chapter 4: In vitro biofilm formation of P. gingivalis.
11. Figure 4.3: Overview of the optimised parameters for in vitro biofilm formation with P. gingivalis.
As can be found in:
▸ Chapter 4: In vitro biofilm formation of P. gingivalis.
- 12. In the definitive edition of the doctoral defense, this is page 106.
13. Table 4.5: Validation of the optimised crystal violet protocol.
As can be found in:
▸ Chapter 4: In vitro biofilm formation of P. gingivalis.
- 14. ▸ http://www.ncbi.nlm...., Mark-Anthony Bray e.a., Advanced Assay Development Guidelines for Image-Based High Content Screening and Analysis, Eli Lilly and Company & the National Center for Advancing Translational Sciences, last revised 1/5/2013.
- 15. ▸ Chapter 5: Optimisation of an in vivo model for P. gingivalis.
- 16. Table 5.1: In vivo oral infection models of periodontitis.
17. Chapter 6: Biofilm formation and DPPIV as virulence factors for P. gingivalis.
This chapter is based on:
▸ http://femspd.oxford..., Sofie Clais e.a., Importance of biofilm formation and dipeptidyl peptidase IV for the pathogenicity of clinical P. gingivalis isolates, Pathogens and disease 70 (3), p. 408-413, 2014.
- 18. ▸ http://www.nature.co..., Song Lin Chua e.a., Dispersed cells represent a distinct stage in the transition from bacterial biofilm to planktonic lifestyles, Nature Communications 5, 21/6/2014.
19. Page 26: Mediators of host immunity.
As can be found in:
▸ Chapter 1: Introduction.
20. Clinical isolates K4 and K5 can be found discussed on p. 147-151 in the printed version, or in the results-section of:
▸ Chapter 6: Biofilm formation and DPPIV as virulence factors for P. gingivalis.
This chapter is based on:
▸ http://femspd.oxford..., Sofie Clais e.a., Importance of biofilm formation and dipeptidyl peptidase IV for the pathogenicity of clinical P. gingivalis isolates, Pathogens and disease 70 (3), p. 408-413, 2014.