Open Source Antibiotics Series 2, 19 Feb 2021

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Open Source Antibiotics Science Update Feb 19 2021

Description

Weekly open project meeting for Open Source Antibiotics Series 2.
Full Project: https://github.com/opensourceantibiotics/Series-2-Diarylimidazoles
Relevant GitHub Issue: https://github.com/opensourceantibiotics/Series-2-Diarylimidazoles/issues/56
On the call: Professor Matthew Todd, Dr Dana Klug, Giada Sabatino, Ireno Demmangngewa, Dr Alex Vaideanu, Prof Andreas Schatzlien (UCL), Dr Chris Swain (Cambridge MedChem Consulting), Dr Lori Ferrins (Northeastern University), Anthony Sama (citizen scientist).

A

All right all right uh afternoon, everybody and welcome to open source antibiotics meeting on the afternoon of friday, the 19th of february. um So we have a bunch of things to talk about today, um including some recent potency data and also talks data. Is there an obvious way around? We should do this. Maybe talks first. Does that make sense to everybody um and then maybe alex can uh present data?

A

If that's all right, do you need you probably need them privileges to do that, uh in which case, let me make you a co-host so that you can do that. Perhaps.

B

I just tried to share, I think it works. Oh.

A

Yeah, does it okay.

B

Can you see my screen.

A

No, yes, yes, yeah.

B

A

B

A

So if you're happy to take it away, let us let us know.

B

Okay, f5 yeah, uh hello, everyone, uh nice to see some faces. I haven't seen before uh thanks for asking me to show you today that you've been waiting for so long so first off, I just wanted to talk a little bit about those response. Curves because I'm going to be showing lots of them and I just want to make sure that everyone is on the same page. So when we do toxicity screens, we generate this type of this type of those response, curves, where we look at the kill or viability.

B

Whichever way around, you want to look at it as a function of concentration and typically these curves take this sigmoidal shape. That's, ideally what we want to see to be able to fit uh the data nicely and the way we fit. The data is by using a mathematical model, described by a four parameter, logistic curve, and that is described by these parameters here. So you have the minimum effect, namely the concentration at which you see the the least circle, uh the maximum effect maximum potency.

B

Then we have this um slope, so we can fit an asymptote to the sigmoidal curve and we can get the slope which is called the hill slope and that's important when comparing compounds of especially thinking about mechanism and then we have something which we call the ec50, and that is the 50 of them measured effect.

B

um That is because so, ideally, this curve would go from zero to a hundred percent, so maximum and minimum would be zero percent. We have a concentration at which we have 100 kill and we have a concentration at which we don't see.

C

B

Impaired growth anymore, um but that's not always the case, so if for for whatever reason, uh the the concentrations that we are screening are not potent enough, and this slope this uh curve starts higher than zero percent. Then we have ec 50 and that describes again percent of the observed effects that we yeah of the observed effects.

B

Now the cc50 uh which is more commonly known, is the ic50 is uh something that the ceo add protocol has looked at and that's why I'm including it here as well, because I guess that's the figure you guys are looking for, and that is um the concentration at which 50 percent of viability is uh reduced. So um typically ic50 is higher than ec50, but it doesn't really make sense to talk about ic50 unless you observe 50 reduction in viability.

B

So the way we conducted the experiment was following the protocol uh supplied by dana, and um I guess the way people do it um for the seo. Add that was in line with the compound solubility, as as dana said, so we started at 32 microgram per milliliter. We diluted two times and we uh observed uh the response for then concentration points down from there. So this is the curve for tamoxifen, which is the uh control they use in the protocol.

B

So this was my first go-to um just to make sure that what we are doing is in line with with them. What was them before so uh otherwise nothing would be, you know, uh wouldn't be building on each other's work, so um you can see easy. 50 is 11.12 microgram per milliliter cc. 50 is also similar and that's, as I said before. Typically ic50 is higher than ac50, but this is within the standard deviation. So that's fine. The value quoted um by the add protocol is nine plus minus two microgram per milliliter.

B

So I think this is good enough.

B

uh Obviously, it's not exactly the same figure, but uh I will also talk about the fact that we used a different assay uh as opposed to the coa dd protocol, to measure the response. uh Now a21, uh quite potent we don't see a 100 kill, but nearly there so cc50 is 1.65 microgram per milliliter, um I'm not sure uh what else I can comment other than giving you the values uh I I'm summarizing again, so maybe that makes more sense so uh 822 again 2.86 um again, we we don't see a 100 kill.

B

So this is not a full response, um but the data seems pretty consistent. 861 is one that I couldn't fit the data because uh yeah it doesn't line up with it with a dose response uh behavior. So I can't calculate an ic50 because we don't observe uh 50, kill so uh 862 same same story.

B

uh We don't observe 50 kills. So I guess this would be one that you don't really you're, not also not interested in uh 865, uh again quite potent and similarly important to eight to two. I guess uh two point: zero. Seven microgram per milliliter, eight six nine is similar to tamoxifen in terms of uh potency, so 13.74 microgram per milliliter.

B

um We do observe fifty percent kill, so uh that's good, uh eight, seventy again quite potent uh less potent than a two two and uh 865, but uh still potent we observe 50, kill uh and a71 slightly less potent than 870, but still potent enough. I guess um so. This is, I guess, the summary um they in the protocol.

B

They classified the compounds as whether they have cc50 lower than 32 microgram per milliliter, which is the case for most of them apart from a61 and 862, and then we also observe um a d, a maximum kill higher than 50 for most of them apart from 861 and 862. So that's in line with the ec50s, the calculated dc50s and ic50s.

B

So, just briefly to comment on the essay that we use because, like I said, as I mentioned in the beginning, we used essentially a different way to detect the same thing. So um the way we measure cell viability is by looking at a mitochondrial activity, um metaboli metabolic activity of the cell, essentially and the way we did it is versus via the mtt assay, which is a uh it, gives you a colorful compound, which you can measure the absorbance of, and you can um associate that with viability.

B

uh The reagent used by the co add protocol is restylane, which in turn is a fluorescent reagent. So um there are some parallels to be drawn there in that the mechanism of action is similar, but obviously florence is more sensitive but, on the other hand, requires more specialized equipment black plates. So overall it would be a more expensive thing to do.

B

Also mtt it in terms of just classifying if ic50 is more or less than 32 microgram per ml. I can just look at this plate and tell you that my ic50 is going to be between these two concentrations. I don't even have to read it, so that's just very high throughput way to screen things if you should be so inclined um again with.

B

As with everything, if you have it wasn't the case here, but because all compounds were colorless, so their spectra didn't interfere at any of these wavelengths that we are reading at either for absorbance or fluorescence.

B

But that is something to consider so for rest as urine, because it's a fluorescent-based technique, we will have to run the assay in a phenol free medium, just to make sure that that fluorescent doesn't add any confounding um and then uh the reason we did mgt is because we're very familiar with it, and usually it works first off as I promised actually. um But it wasn't the case for if there were quite a lot of things to optimize with this particular cell line, because they it isn't a dear end.

B

So from that point of view, probably wrestling is a more practical way to do it and um the the standard deviation and generally, the variability of the data would be less because you also have less washing and removing of reagents and adding things um which is fine for shorter period essays like this, but it's probably not uh suitable for longer term. If that was something you were interested in, so just to show that I also uh did the resazurin for myself just to check that we um again, the cells are similar and everything is consistent.

B

So here I'm showing the curve uh this time this this readout is the fluorescence readout. So the cc50 that I calculate is 19.74 microgram per milliliter. While again, the quoted value for in the add protocol is uh nine plus minus two. So again, this is within well the next order or the next concentration point or within the same, but still within the same order magnitude. So I think this is uh appropriate.

B

So, finally, uh things that I would like to do with this data is just to look a little bit more at the variability and understand where it's coming from so do some more comprehensive statistical analysis and also group by up operator, because a lot of this data that I showed you is an aggregate of both mine and reno's uh experiments. So I think um discerning by that confounding variable might show different patterns and then also transfer this analysis to r so that it's reproducible at least computationally.

B

And finally, I just want to thank reno because I'm showing you the data, but it was very much a team effort and thank you for listening, okay and sorry for going five minutes off. Oh.

A

Don't worry, that's fantastic! Thank you. Any immediate questions. While the slides are up.

D

It looks like you have different hill slopes for different compounds. Yeah. Is that an artifact or is that real? Do you think.

B

um I I don't think I can comment on that without knowing any more about the compounds.

B

I think the the hill slope is good to look at if you have similar, ic50s uh or ec50s, uh but then obviously also because this this screen is on the same kind of concentration uh pattern in that we did two times dilution.

B

Obviously, if you have compounds that are more soluble and more potent, then you can start at a much higher concentration and then do different types of dilution. So then that slope will be different, so it I think it's the hill slope is probably something to uh look at when you know something about the mechanism and you have similar ic50s or when you want to understand. If uh assays done, maybe slightly differently with different concentration regions, you can again build on that and compare.

B

As far as I know, okay does that answer your question.

D

To some extent, yes,.

A

um So the uh I mean it's the obviously we're seeing some fairly high toxicity with some of these compounds, and what we want to know is how that tracks with potency.

A

um That is against mrsa, because the two compounds that look like they have lower levels of toxicity right, six, one and eight six: two: um does anybody, so those would be the um yeah slides, yeah, six and seven they they appear to be lower than that. So can.

B

I ask: are you in? I thought I thought you were interested in things that are toxic, so I got it. The.

A

Opposite, so these compounds are intended for treatment of um bacteria. Okay, um this obviously doesn't change your numbers, uh the um no so we're interested in compounds that are, you know, effective against mrsa and and not toxic here, um so asus 1862 with a murder that gets the least toxic. Looking of.

D

Also not active, I think.

A

D

I'm not active okay.

A

So I wasn't that's the thing we've got to do is correspond, those with potency against. Does anybody happen to have that to hand or.

C

Yeah, it's the three piritol and the two and the four pyrital that we're replacing the two uh on the I guess southwest we're calling it um so you know what I'm talking about the green.

E

C

Yeah, it's on github issue, number three as the structures of all of them, so those okay yeah, those aren't- are not active.

A

Could I um could I just rest back control from the alex? That's a screenshot.

A

Thank you, so the um third issue number do you want to share it or if you got it right there.

A

Right, okay, so yeah the so x, 1 x, 6 2 are those two crucial compounds that were synthesized to try and interrogate whether 821 is is effective, because if it's too paradile- and it suddenly looks like that's the case right, so the at1 is definitely toxic in a way that 861 and 862 are not.

C

um Right, I guess 869 is interesting if I bring that correctly, that compound is not active against mrsa, but I think alex it probably was showing some toxicity. So that is interesting and it might suggest that this mechanism, we are worried with the metal chelation, is not the mechanism of toxicity.

A

Right, it's lower 869 right, it was like 13 or something yeah. So.

B

821 is the most toxic and then 869 is the least toxic.

C

Aside from eight six one and eight six, two right: yeah yeah, okay,.

F

Can we can I ask um the the um differences between the compounds alex? um Are we talking orders of magnitude or what are we.

B

Doing I mean so it's one two, three five thirteen and then two that I couldn't calculate.

F

I mean um by by the standards of of normal cytotoxic compounds, that's not a huge difference and um I'm I'm wondering how much one should read into these. uh You know um whether these distinction are fine enough for you to come to conclusions about mechanism and and so on, so so um and and alex you you uh I mean if you were going to do a statistical comparison of of what is significant and not so significant. um I think um it may shed more light on how different they actually are.

F

You know and then um can I ask what sort of toxicity do you expect.

A

If there's no, I think there's no expectation, it's more, that we hope below so.

B

Can I ask, why did they use tamoxifen then, as a.

A

Oh this is they use that, for coad? Is that right is that yes, yeah.

C

A

I don't know why they use that. That's just I mean they're a high throughput. You know screening facility versus the escape pathogens. I don't know why that one.

B

It's important for mechanism.

A

So so mechanistically the compounds- where do you suspect, expect them to act? We we we're trying to figure that out. um So we we've had some experiments done um at unc, chapel hill to try and elucidate that um they came.

C

From um kinase inhibitors, so that's potentially a mechanism.

F

I'm just asking because of um these um essays were um doing a relatively short term toxicity or, um if I mean alex, will correct me if I'm wrong, but um one of one of the concerns you might have, whether you pick up all the toxicity in the sense that if you have compounds which would interfere with dna replication, things like uh that are linked to the cell cycle.

F

um What would all of those cells have had a chance to cycle and therefore notice that they've been poisoned or um not? So um you might get a I mean it won't get less toxic. I suppose it's the bad news.

F

um You might find more subtle differences um coming out and I suppose, um as alex alluded to this, this essays is you know why non-ideal just from a number of mechanistic elements of just how it's done and um which, which was one of the reasons why all this took a bit longer and what one could consider um just doing a an mtt in a conventional way and see what difference comes up there and whether those are I mean, I guess we have some indication of where uh ic, 50s or easy 50s would sit, and we could um probably look to optimize um the range that we cover um to to um fall within to that ic50 range and and um we can see whether we can uh in that way get a more a tighter difference between the compounds.

F

But having said that, I mean um if it's yeah, I'm wondering how how much differential we can um see here and how much we can read into. That is what I'm wondering about.

A

I know what you mean- I guess all along we've been concerned that the the toxicity is associated with um this two-paradile motif, so the the things we were just looking at up here were the compounds that we sent to you, um and it's always been this.

A

This motif, that's been the issue or you know the worry, I suppose, um because potency falls off if you against bacteria, if you remove that two, which is having the nitrogen here, so these compounds are not active against mrsa, and so we've always thought that maybe this is it and that maybe this is going to be something more generically toxic, which is the whole reason for the measurements.

A

So it's more about whether we're seeing um a correspondence between potency uh against mrsa and and toxicity and and in in these results we we kind of are in the sense that these are definitely inactive compounds against bacteria, and they are the least toxic that we're seeing here. So it's a sort of broad class of toxicity that, rather than a neat trend, there are some other compounds here, though, that came from laurie who's on the um call um and again mo. I don't know if you can see.

A

Hopefully you can see these that they're they're um thick enough, but the um uh so in these cases we we've again got um it's a slightly different measurement and lori will be able to tell us. You know what the difference is between these and the kind of measurements we were just seeing.

A

But again, many of these comments have the two paradigm motif um and- and you know the one that doesn't eight five five is the one with with the with the least toxicity, um but we were a little bit heartened by some compounds that that had the the motif so um eight five six here has a two paradigm motif and didn't look particularly toxic, but was inactive, and so I guess we were interested previously in this compound 983, which is this guy up here, which has the motif that we're worried about, but has this kind of tertiary nitrogen um in this position with an alkyl group on it here, 983 wasn't particularly toxic, so we were kind of hoping that this would have activity.

A

This may be the segway into the next bit right because we now have potency data back on these am I am I have I got a crosswire. Is that right? That's correct! That's right! All right! So 983 is the crucial one. So should we just quickly deviate over to those new results as well, um so that we can see if this holds, because 93 is the crucial one, um whether we have an example of a compound that is uh not particularly toxic according to that measure and uh okay.

A

So these are they right? You just put these up. um So what was I saying? Eight five, three. What did I say? What was the number I just said. Sorry.

E

It's 983 and it's lowered down no.

A

But these are the sorry. These are the wrong compounds. Right. Yeah, that's right blocks, got it sorry, wrong!

A

Yeah, there you go.

E

Yeah, that's it 983.

A

Which is potent as potent as we've had that guy?

A

So these are the new results, um just posted from paul um on laura's components,.

G

Or you know showing some activity, which is great, which is actually really nice, because we can just tweak that to infinity right. You know.

D

I think you want to put a few more of these through the toxicity essay to show that you have got a robust distinction.

A

I mean 983 being key all right, yeah.

G

A

G

Yeah, because that also has vre activity as well 97 the cyclopropyl.

A

um There were, incidentally, while we're looking at this, the there were the mssa results, which is, um as I understand it, uh the same but lacking efflux pumps, um which paul also looked at, and the potency tracks pretty much exactly so. The conclusion from those data would be that eflux is not our problem.

A

In other words, it's not the reason for the limitation on potency.

G

The fact that entryway always says they pass regardless of potency.

A

Right sorry, entryway said they all passed.

C

No, a pass and entryway is meeting all three criteria. These all meet two out of three. They never have the primary amine, that's the third. So I think that, actually that's not a pass.

A

But experimentally this means they are passed right because they're not being pumped out.

D

It means they're, not substrates, for that pump.

F

Yeah, I was wondering to ask those pumps: how do they in their substrate, specifically specificity, relate to mammalian cell pumps.

A

A

Sorry, how do you me so the the entryway are uh referring to the effects from the gram negative bacteria right?

A

That's the those are the so you you we we were wondering if the the flux pump and mrsa pumping them out is that are you asking about something else.

F

Well, I'm asking whether whether um pumping out um in bacteria and pumping out in cells is in any way related in terms of substrates or whether whether differential pump activity can bring differential.

F

You know if um something can be pumped up by mammalian cells, but not by bacterial cells, whether that is a root to some specificity.

A

I don't know the answer that question, I'm afraid.

A

A

I mean so just looking at these.

A

um These latest data- I guess in terms of as you were just mentioning, alex future work, then, given the broad classification we just seen in the results, you have of you- know low cytotoxicity, corresponding with low activity against mrsa.

A

These new compounds at the top here would be if there was any more work that could be done, and that depends on bandwidth and resources and everything then uh checking out the toxicity of these compounds, which we have in the lab would be of a high priority to see if we can find the therapeutic window there. A selectivity index.

G

Window I mean it's possible to just: could we just run 979 and uh 983, because those were the stars of the show just run those two.

A

uh Yeah, and also I mean yeah- I mean two or three I would I would guess just if we can have something which is inactive and something which is active, they're all pretty active right, but yeah we're trying to cover the spectrum. I suppose um just going back yeah. Do you.

B

Have any golden compounds or like something so tamoxifen, because I thought tamoxifen is important for mechanisms. So that's why I wanted to make sure that that's all in line, but do you have anything that is commercially available or something that we can benchmark? I think that would be useful.

D

Yeah, unfortunately, we don't know the mechanism of the toxicity. Otherwise you know we, we could suggest a positive control.

A

I mean uh yeah, but I mean I think that the whole point about the he isn't the same company right was that we would get the similar. We would have confidence that, even if there were assay differences, that uh yeah.

B

Yeah, no, no that, but I mean in terms yeah. I guess, if you wanted to be able to discern better, we.

G

Don't have any information on the moa or mlt for that matter.

F

I mean there are two compounds we looked at are quite different now, top and bottom: isn't it structurally um well, they have the motifs that you're interested in, but otherwise they're really quite different. Isn't it well.

A

They have this different. They have this link here which, which we don't initially, we didn't initially have yeah.

C

A

Kind of compound this is our standard 821, so yeah they've got this little bent core here, but the rest of it is broadly similar. um You're worried that we we may be of mechanism hopped.

F

Yeah so so um yeah essentially, but I mean um I don't know, don't know it's not like yeah.

D

But does it matter matt if we've got a new mechanism as long as they they they're antibiotics, yeah.

A

Phenotypic survey forever yeah.

F

Yeah I mean, I guess what I wasn't sure about is um in terms of trying to discern toxicity of particular moieties compared to others. That's it's quite a big jump. Isn't it from from bottom. To top I mean.

G

We could could we not try making eight nine three but lop off the bottom. Two period, aisle.

A

Yeah I mean I I would want to. I would want to just double check these first, um so yeah so 980 is is fairly toxic according to this assay, um which is you know remarkably, similar to the one we're looking at with different alkyl chain on here I mean I definitely want to get these checked out again.

E

I think that one so 980 actually has a methoxy on the I think on the imidazorine core yeah uh upper position, yeah yeah. I think I think the.

C

Structure's wrong daniel.

E

Is just going to die.

D

Sorry, yeah yeah.

E

Sorry guys, sorry, oh.

A

That's all right there. It is yeah right, yeah and.

E

So we have seen we have seen differential toxicity. I think that that's the position that's norm like normally that one is better than the substitution pattern on 855, but I would need to go back like there's a lot change between those two.

H

Good news, it doesn't seem like we need a substitute substituted core.

E

No, no, you don't well. It doesn't look like that from this data. Okay,.

A

Sorry yeah! No, so then it's 981 98293, because that's the jump tenfold.

E

Yeah and so we have um pmm data on 983 as well on all of these and there's also a big jump in terms of pmm data on 983, it's 23 micro, molar versus single digit and animal or sorry single digit micromolar. For the rest.

A

Okay, I mean it's really interesting that you got these differences with these tiny.

E

Yeah, well, I find the difference between 983 and 987. Sorry 979, quite striking.

F

Well, I mean one where you, where you just um showed a tenfold jump. um If you can talk me through that, one again, yeah, so yeah you're going here or nine eight one, nine eight, two, nine eight three.

A

Yeah so you're going from five to fifty.

E

I don't have it, I don't have an explanation for you in terms of that one I mean, we also see it um like that's what I'm saying. We also say it in a pmm cell line, and I.

A

Can't explain it so andreas yes, structurally! It's just a modification of this outcome group so and which one was the more toxic one. These are the more yeah.

F

This is the solubility is not an issue there.

E

um It's not it's, not amazing um matt. Would you mind if I share my screen for a second yeah.

A

E

So I added um the host cell line. Sorry, not wholesaling uh the the toxicity cell lines that we have as well as uh add me, data for these compounds and 6886. Sorry 983 um is definitely the pick of the bunch. In my personal opinion, um when you consider post cell toxicity, um but no I mean it's.

F

E

uh This is measured at ph, 7.4, okay,.

A

And the solubilities measure.

E

Yes, it's uh thermodynamic solubility.

A

I mean it's not the worst. 983 is not the worst, but it's not grey, but it's not the world. It's not.

E

Great yeah, so we have other analogues at the moment. uh The chemistry equilibrium is actually making some other derivatives where we've kind of got um primary alcohol on sort of the the alkyl chain. So they could actually be quite interesting. I'm hoping we're hoping that that will help in terms of solubility they'll, probably give us another.

G

Thing that I was talking about was uh amide or amide functionality. We start with a uh acyl chloride and then attach it to it.

G

Start with, like chlorobutanoic acid or something and then attach the chains I'll draw up a couple ideas. I've had, if that makes sense,.

A

um Just I mean if the implication is from these combined that we should be pursuing this core more because there has a better chance of selectivity, but we want to get them checked. Yeah.

G

I agree totally, I think uh the the five plus six core is probably gonna be easier on all of us and probably easier to make to. We do need to verify before we hop.

A

Because this is interesting because of course it comes back to what we were talking about last time with lee um and the mechanism of action studies that if we had things that were generically toxic, then it's less interesting for him to do the experiment on because he's using mrsa lysate- um and you know maybe we're using compounds that are just generically toxic and so well. I guess it doesn't make that much difference. But if we had a compound here that we were, we were more happy about the selectivity index.

A

For then that might give us more meaningful data for the mechanism of action, stuff.

D

I think that's absolutely correct.

A

So if 983 holds up, then that will be a better compound to use for. The repeat of those experiments is that that makes sense.

E

And as we're making more of these derivatives, we can certainly include a included like set aside a sample to ship to you as well. If you're interested.

A

Cool I mean yeah, we don't I mean we don't want to. You know, interfere with whatever you're doing, but if there's common chemical space that we can work on yeah, it is yeah.

E

The this this is kind of the the area that we're pursuing at the moment and then we're also substituting off the paradigm as well.

A

Very interesting: well, that's very useful. um I mean yeah it does. It is important that we try and um prove that so we have sample left right from you laurie in our live danny we've got still got sample from laura left right now we didn't consume all those donated samples for the.

C

uh Well, I did dissolve it all for paul, but I don't think he uses all of it in his essay, so I can check with him and to see if he has some stock solution. Left was.

A

That dmso yeah, okay, alex. How do you do the asset? Do you take dmso samples.

B

So I I dissolved everything in the msi at 3.2, milligram per ml and dilute 100 times from that, and that's the starting point of the curve. Okay,.

A

Okay, I mean so you're you're busy with everything else, but um can we put our request in in your your bandwidth queue to see if we can have one or two compounds extra measured, yeah.

B

I I I think we know now better what we're doing so. It shouldn't be that big of a yeah.

F

I'm wondering um do you know how how tedious is the assay compared with the with the working with it here in cell lines, and because I mean at the moment it seems. The only reason for this assay in this form is that it compares to something someone else has done elsewhere.

C

F

No mechanistic underlying reason that that says this is showing either in the cell line or um how you do the essay. You know. If you wanted to compare things directly, you could you could choose any sort of toxicity asset you would like to use. Isn't it.

D

Yeah, I think that's right. I think all you need to do is, from our point of view, is to take a handful of these and do a sort of bridging one where you do them, both in both assays and show that you get parallel results.

F

And um yeah I mean just just from my very limited experience with with trying to develop um anti-material agents and often a sort of uh one point where we've seen uh them acting on is, is sort of interfering or binding dna and and not necessarily super specific and and so I'm wondering whether just um sort of doing a bridging.

F

With with some of the compounds where we looked at, where we, where we just now um looked at these pyridine um segments and and then your others, and and putting that into a mtt that runs over the the seven days alex and and it may, it may show more differential between the compounds. And, if that's the case then maybe gives you better guidance of of where to go.

B

And which cell line.

F

Well, I mean you could could be any. I suppose, really it doesn't matter too much I mean um adherent. Obviously I would go for.

B

Yeah I mean I just mean it should be human right, because otherwise yeah.

F

Yeah yeah well, probably doesn't make the difference much different with mammalian but yeah. Okay, if you have more.

A

Justification to do.

F

A

To use mammalian so yes, what's.

F

Your what's your.

A

Preferred assay that would be operationally simpler.

F

Well, I mean it comes. Obviously: typically we look at cytotoxicity trying to find it um and and um we then, when we have it, then you would maybe look. You look at compare different cell lines which have a different genetic makeup, and then that gives you a hint about where you see more toxicity, and so we don't have a particular preference. Anything that's easy to work with for initial screen is fine, I would say um so, if I'm not mistaken, you're looking for cancer drugs.

F

Typically, yes- and um I mean I don't know whether we've got the fibroblasts kicking around wi-30, we used to have those.

B

Yeah we do have them.

G

Fibroblasts would be good because uh most mrsa infections start through the skin.

F

B

It's not something. I've worked with a lot, but yeah.

F

um They should should be very well behaved. Actually so shouldn't make much difference, um or I mean if it's quick, uh we could use any ones. I mean.

C

F

You know there's this uh direct comparison between compounds that you I don't know if you have a short list of six or something, and you want to see how they all compare in the same essay, then that's that's easy. If, if continuity is more important, then I think um working with the what about changing to in still still a pain, isn't it yeah? I say.

B

uh You you nervous yeah.

F

B

I mean no, we can do it just yeah again, we can do it, but as.

F

You spend a lot of time actually on it. Isn't it uh comparatively to mtt.

B

I mean I don't know how, if we want to do the week-long protocol, where we try to discern like some more subtle differences. I I don't know how, because obviously fluorescence is different to read out so then we have to again optimize cell number.

B

F

I mean so so the question starts in you know why, with a question of how discerning is this essence is, is discerning enough to want to draw uh conclusions about what structures might carry um toxicity, and if that's for those compounds that we've looked at something you still want to pursue, then it might make sense to do another assay, slightly different form with those compounds.

F

If, on the other hand, um your thinking is now that um well, maybe this other structure with the uh floor and and the different length alkyne chain is to be explored. Well, it's a slightly different scenario. I think- and it comes a bit down to your priorities in this.

A

I think the first priority is to identify whether we have a selectivity index, so we have compounds that are potent against mercer and not cytotoxic.

A

If that's the case, uh if that's not the case, then we have to abandon ship, publish and get out if it is the case, um and we can explore a compound that gives us that index then yeah. The next thing is to say: well, it will be great to know what the mechanism of action is.

A

um Intellectually, you want to know what it is for sure, but just being a little bit pragmatic, you don't want to pursue a series which has no future um if you, if you can avoid it, um but but if it is the case that we have a selectivity index, um then we definitely want to know what the mechanism action is and if there are clues that can be derived from what kind of toxicity we're seeing. That would be extremely useful.

A

F

I mean when I look at those numbers, my eyes, glaze over very easily there's different ic, 50s and easy 50s, and an order of magnitude is easily lost on me, and um I was wondering whether you've tried to um put this graphically in a way where you, where you sort of see a correlation between both- and I personally would find it much easier to um assess. But um that's just me not being great with numbers, I suppose, but no.

A

Sure that would be. That would be. I mean, I think, that's what we're all doing in our minds is creating that graphic on the fly um yeah normally normally I I come a cropper doing that, um but I mean at the moment we do then have data from three different toxicity assays slightly different toxicities, so lori's essay is slightly different from the co-ads one, and we've got another one again introducing a fourth one.

A

We we have to be really sure that we we have those bridging compounds that have been tested in all assays to make sure that we want yeah. The one diagram can can can take all of those data points and we're seeing something: that's real, but how.

F

I mean those different essays. um How different are the numbers that come out? I mean, as I said, I don't have a good feel for it, but you know with with um cytotoxicity under cancer compounds, I mean we normally do one in ten dilutions and um if, if it isn't ten times more, it's basically the same. You know this is the how discerning I think these essays are um and um yeah.

F

So that's so that maybe if you could argue that, if, if um they are sort of in the same ballpark, a similar ballpark could still at least to to get a feel for things, because, right, as I said, I find it difficult to get a feel for things by just looking at the numbers.

F

Put put it all together and see what you get I mean if, on the other hand, the differences are big. Well, I guess it's good in the sense that it's it's discerning, but I just don't have the feel I don't you know. I can't pick that up that quickly.

A

I think um I mean again the number that people want to see in this area would be that selectivity index, so so the the toxicity difference between mrsa and and mammalian um and you'd expect to want to see about a 50-fold difference if you're going to have something of value. So you know one microgram per ml, toxicity versus a 50.

F

A

Sorry, one microgram of potency versus a for the sort of 50 something for mammalian. You know you want those kinds of numbers I think, and if they get much closer than that you're gonna be you're gonna be nervous.

A

D

I think at the moment, though, you just want some evidence that the the two things are divergent yeah.

A

That's right: that's right!.

F

Yes and um do do you over the top of your head know which which compounds are most divergent? I mean because there's there's absolute activity and then there's divergence, and I mean I I just by looking at these numbers. I cannot get that so quickly in my head. What's what you know and um if you're looking for structural clues, you you might, um you know, try and create a sort of ad hoc index like that and maybe see whether that um structurally drive guides you in a different direction from just looking at it.

F

A

Well, the best was the with the compounds we were just looking at. I think those had the best. You know the biggest difference between the two potencies yeah.

D

Yeah, I think we need to test one or two more of those and just see whether ben's a middle structure.

G

Hop is in the pipeline right.

A

Yep um sure yeah yeah, so there there's some. We don't have time to talk about it now. I think because we're coming up on the hour, but there is chemistry in the pipeline that also diverges a little bit more too yeah.

A

Okay. Well, it's been very useful, um so I guess I'm really good again.

D

I'm saying really good.

A

Yeah yeah, so I mean first of all, thank you so much for getting the data absolutely crucial for us. So that's really been very helpful um and thanks also in his absence to paul for getting the data back on the compound. So we only gave him like three days ago, right two days ago, so yeah his essay's fast to get to get the potency on against the bugs, um and- and I guess we have an action item to follow up with with you guys on evaluating a few more compounds.

A

So we can be sure what we're doing and those are going to be pretty key experiments. We haven't had time to talk about it today, but yeah I mean the chemistry is in the works and um giada has been looking at mechanisms of action stuff as well, and there's a there's.

A

A github issue on that for uh she's been trying to find structures of um pdb files crystal structures of those that list of proteins that that lee grace gave us last time, she's been trying to find crystal structures of those potentially with things bound to give us a sense of whether or not things that are bound may be similar in appearance to the molecules we're looking at.

A

You know whether there's any clues and the literature for for ways in which our compound may be interacting with proteins, so she's uh beavering away on that, which is great, um but we haven't got time to talk about it. I'm afraid I don't think so. I have to say that till next time was there anything.

C

A

B

May just ask: can I ask a question sure.

A

Yeah, please sorry.

B

um So uh because you mentioned that you have data well, I'm guessing similar data toxicity data on different cell lines. Is that kind of? Is it possible for me to see that or is like just so to get an understanding or.

A

B

Did they misunderstand that.

A

Toxicity on different cell lines.

C

Yeah I'll send you alex okay, yeah.

F

Great thank you. Well, nice meeting. Everyone yeah.

A

Thanks for coming along, you know always welcome alex yeah and.

G

A

Everybody next time, all right thanks, so much bye, bye.

B

Bye. Thank you. Bye.

C