[00:00:01] >> And I want to talk about the rational design really of drug carrier sender equally of political will to contrast agent carrier solves some before I get much further talk about the people who are really doing the work as opposed to talking about it. I have a six or seven stupid sixty and so right now at C.S.U. and with Dr Allen Kaunda who's just been hired by disappointment. [00:00:29] I share three students. I've invested nice group and I'm not really going to talk about this work today at heart valve evolution and with every Kaufman's group of the University of Iowa and there's a long list that goes below this but these are my primary collaborator son. Ken Beck is a radiologist and Jeff McLennan his opponent ologist interests in my group are are quite wide. [00:00:57] I have a extensive project on computation of fluid dynamics simulations off to long and what we do here is we take C.T. scan images off the lung geometry and perform simulations on those with the idea being that geometry is actually critical for the delivery of inhaled drugs to the lung. [00:01:22] We've just published this work in A.B.M. me showing that the geometry is actually the critical determining factor and the proposal that we have is to be able to do patient specific C.T. scans or some other imaging modality and use that to optimize drug delivery for a given patient. [00:01:47] So clinically now you see up to one hundred percent variation all else remaining the same simply based on geometry and we hope to be able to reduce that interest for Torrie drug delivery another section of. My group and both these things. I'm not really going to talk about in any detail today I'm just giving you a quick overview now we look at fabricating controlled and modulator drill East particles for delivery of drugs to the lung and the target that we have here is for insulin for example to be able to establish a weeklong supply of insulin in the lung and then trigger the release as needed for systemic release into the body and the way we do this is we have nano particles which we crosslink to form these large cross-link to glamorous and these are the insulin or whatever drug of choice is contained within each of the nano particles and then you expose internal surface area by introducing a cleaving agent. [00:02:49] We've done this for insulin rhythms for superfluous Assen. But it's a topic for a different seminar altogether. What I'm going to talk about today are what I have down here. We've applied many of the drug delivery concepts that we have working to the delivery of contrast agents it turns out that in C.T. contrast for example the last significant development was thirty years ago with the development of non ionic contrast agents there are huge sharp comings to their use and double discuss that and I'll talk about what we're doing to improve the delivery of contrast agents this work was just recently accepted in academic radiology it should come out in the next couple of months and I'll talk about targeting of these carriers using Milligan based targeting technologies that we've also been developing in my lab in collaboration with Ravi belum Congress. [00:03:47] OK what are the issues today with. I had a neighbor to contrast. One of the big problems is. It turns out that clinically relevant. Contrast require something on the order of six hundred milligrams per kilogram of fire and I again by itself elementally is highly toxic so you have to work with some compound off the ID. [00:04:17] It does need to be highly soluble in a quest medium in order to be able to image the blood pool for example. And by the time you get this much Iranian into solution the the of the solution is huge and is going to cause all kinds of other problems when you introduce it into the blood. [00:04:38] One approach is to simply reduce the total number of I guess of molecules that are carrying the atoms and what people have done is go to excuse me non ionic materials Wes's ionic materials and the idea of being the nominee materials do not dissociate so one molecule that contains IO Dean is going to remain as one molecule and therefore the US without the would potentially be fifty percent of the case where dissociated into two species. [00:05:14] And a classic example of that is Io hacksaw of which we actually use in a lot of work and it has fifty percent of the osmolarity of a similar compound tries so it. The second approach has been to simply increase the number of fired in atoms per molecule and there's been a relatively recent development visit pake which has six in Adams versus three in the case of fire Exxon and I think the next couple slides I showed their structures Yeah OK so here is Io hacksaw again it is a non ionic contrast agent it's got three iodine atoms per molecule versus visit pick the newer material which has six in atoms per mile. [00:06:00] So conceptually if you got these both would say Mauler solubility you would have twice the number of fired in atoms in the solution versus this at the same ospital ality OK So that's one approach that people have used the other approach there used is look at atoms other than I or D. in as X. ray absorbing agents gadolinium and Dysprosium are two examples and what I've shown here is essentially the absorption coefficient getting close to the carriage as a function of energy and of plotted here ninety Cavies to about one hundred fifty K. vs which is pretty normal for a C.T. scan instruments. [00:06:45] And what you can see is there is about one and a half to two times depending on how you look at it increases in the absorption over in the case of gadolinium and this procedure mentality and Dysprosium have advantages because they could potentially also be used for AM are appropriately applied. [00:07:10] That's another approach that people have used in all these cases however there is a timing constraint and what I've shown here this is data from Bayes group at Washington University all these small molecules are going to clear very rapidly from the circulation via the kidneys and what I'm showing here is the enhancement in Hans field units as a function of time post injection for a typical I donated compound I think this was actually I accelerate their use and what you're seeing here is a very rapid rise in the contrast up to about three hundred or so. [00:07:54] Hans field units and a rapid drop off down to virtually a nice levels. And this is using a injection in the superior V. in a cave us if you wanted to image the heart for example you would inject this material immediately upstream of the R.C.D. have to do a catheterization procedure inject the material directly upstream of the heart and then the peak and handsome and you get is only about fifty or so seconds wide and if you want it and even within this fifty seconds this is a very rapid rise and fall. [00:08:28] If you want stable contrast you are essentially relegated to nice. Now there are ways of making this better by mechanical means and one approach has been to simply slow down the injection Earlier I talked about a five mil per second injection. If you slowed down to two and a half mills per second. [00:08:47] Then you can broaden and smear out that peak a little bit but this is still a very very difficult image acquisition procedure to do because you've got to time the image acquisition to this peak and so danger action is performed I don't know how many if you're familiar with this part of the high pressure compressed air type injector. [00:09:10] A certain percentage of these patients end up having. Failures in the injector and blowing out blood vessel where you're injecting it. It's a risk frac procedure. Now you can you can slow this down further by doing by physic injection and so we're still talking about one hundred seconds or so and now of contrast. [00:09:37] So what we'd really like to be able to do is take this to the next level and see if we can extend this period quite significantly. Some of the other problems that go on with with the current C.T. contrast agents in order to get these high contrast levels with high concentrations you end up with very very high of a Scot. [00:10:01] So now you have to inject at high speed in bolus in a catheter something that has very high viscosity. And these things clear by the kidneys this toxicity associated with that high speed data acquisition that I talked about and it's very difficult to time scans during that peak enhanced meant so you synchronize the scanner with the with the bolus injector. [00:10:26] And sizable percentage of the time you just you miss the peak and you don't image the blood vessels that you want. Many times you have to work with multiple men over so you have to inject repeatedly and the Iranian loads that you're injecting into the body are huge and as a result you contrarily use these techniques if you've got renal dysfunction already or if you're diabetic the material is contraindicated So we'd like to get to a state just to cut to the chase where we have a T. half for a clearance that is greater than six hours and trying to limit a lot of the toxicity problems we want to see if we can clear the stuff differently and still via the kidneys and we can do that and we want to look at targeting and so on. [00:11:14] OK so how do we go about this. Well it turns out that one way or the other a lot of things in my lab end up dealing with life zones. Because they're very versatile and everyone in my lab can make life zones in their sleep for those of you who are not life as some afficionados they consist of a by layer bonded sickle and the molecules in the by layer are typically some mixture of limpid which have a long fatty acid chain and a polar head group have shown here a fossil saddle at another mean example. [00:11:51] Typically the lab results that we work with also have thirty to forty percent cholesterol in the by layer. For multiple reasons. They increase the rigidity of the by layer and also if you go much below about ten percent cholesterol when you inject these things into the bloodstream to cause him Allah says because they extract the cholesterol from the from red blood cells. [00:12:13] You can make these things in many different structures. You can start them out as multiple layers best circles and then through a procedure that I'll show you later on we actually make these things that are small metal or VESA cols OK when life goes with first conceived off people thought of them as replacements for cells and the idea was that if you injected these things into the bloodstream did stick around for a long time. [00:12:41] Tried it didn't work because it turns out that they're simply not hydrophilic enough and they are very rapidly optimized and they're cleared this is a relatively old quote I kind of like it which is why I put it up. It turns out that you have to increase the Hydra felicity of the surface of the lipids omes in order to enhance their circulation half life and I love this line to be invisible. [00:13:08] One must look like water so if you're in the body that's what you need to do in order to be invisible. So these are called stealth light presumes this is a relatively old technology and this is the basis of a company very used to work and we've put a couple products out on the market based on this using dots Aruba's and chemotherapeutic agent and so the way these work is you've got to buy a layer Boehner Jessica and now with a polyethylene glycol want to kill hanging off the side and this imparts sufficient Hydra Felicity that you greatly increase the blood circulation time as an example of that here is Doc's Aruba's and data that we collected a long time ago he was free docs a rube. [00:13:51] So this is the free drug injected into rabbits. They both start out at exactly the same level and you can see the free drug. It's very rapidly clear that this is also really cleared by the way whereas in self lie presumes the clearance is much slower and you can see even forty eight hours after injection the stuff still hanging around clearly from a bioavailability perspective you can see how this would be FOSS appear ear to the free drug provided it also got to the tumor center turns out it does. [00:14:24] There is a dependence of the clearance on size of the light presumes he has hundred nanometer life presumes this is two hundred four hundred nanometers some time following injection and you can see very clearly that the smaller These particles are the longer they take to clear the trade off that you have is the smaller these particles can't the more liquid you need per unit volume of encapsulate because the surface area to volume ratio considerations. [00:14:53] So you have to trade that off and you pick the particles that are the right size. OK So in that introduction to life was ohms. And this is how you make them yourself assemble limpid in a Chris medium you exclude them through these polycarbonate filters. These are nuclear pour filters which have cylindrical poor so they are not your typical dept filter configuration but they have cylindrical pores and they essentially act as extrusion dyes and you can make a fairly tight distribution of sizes. [00:15:30] All right so now here is our data on IO hacksaw So this is an Iranian compound that we have encapsulated in stealth like a zombie put this in a rabbit its thirty makes per million fired in as those the material that we that be made. We split the injection up actually into two injections that were fifteen or thirty seconds apart so it doesn't really show up on this plot but you'll see the data later on. [00:15:57] Immediately after injection the. X. ray absorption in the blood so this is a Palmeri artery and I'm calling this liver card text but it's actually the hit to have Pratik artery jumps up to about two hundred Huntsville units and it stays flat for the three and a half hours that we ran this experiment at this point our radiologist friends got bored and they went home because nothing was changing after that there's a transience rise in the kidney core and this is primarily because our formulation had actually leaked a little bit before we even injected it. [00:16:38] And so the leaked compound behaved like free irony in it and you correct this for the kidney volume it turns out it's about ten percent which is almost exactly what had leaked before we injected this. And in less profuse tissues you don't get nearly as much of an increase. [00:16:56] That's because there isn't enough blood here for us to see a significant increase in the contrast. OK so this one is without contrast this is with contrast and you can see all kinds of details in here this is a very coarse image I'll show you a lot more details later on but you can start seeing the heart you can see cataract hoods and jugglers and so on. [00:17:20] And it turns out you can also start seeing vascular ization off the bowel and so anywhere where the blood goes and by the way this is two and a half hours after injection of the contrast agent we can see the blood vessels very very nicely. OK so here is a cross-section through the heart of that rabbit. [00:17:45] This is before contrast injection you can see much of this is after the injection of the first dose of contrast and you can start beginning to see some features developing in the heart. Here's after the second those have. Injection and now things are lit up very nicely you can start seeing various heart chambers you can actually start seeing even a coronary artery and I'll show you that a little later on. [00:18:11] And here is two and a half hours later the rabbit moved a little bit but the image is static it's exactly the same as it was when we first injected the contrast agent. Every time I show this to a radiologist their jaws drop and they say how did you do that and the answer is that the stuff just sticks around long enough in the blood stream. [00:18:33] And here is the liver before injection of the contrast injection of the first dose of the contrasts and you can start seeing quite a few blood vessels second dose of contrast and now of things that are really very well lit up and again two and a half hours later things are still it up just fine. [00:18:54] Here is image of the heart. This is the volume rendered image of the heart chambers before contrast obviously you can't see anything after a contrast you can see the error in the poll narry artery left ventricle right ventricle inferior vena cave. You can even start seeing value or details and I'll show you a tick slab rendering that will pick up the Val details Faria OK here it turns out in our rabbit Model one hundred percent of the time we can pick up the coronary artery not a big deal. [00:19:28] Whereas traditionally told in humans with bolus tracking the injection you're lucky if you pick up the coronary artery fifteen or twenty percent of the time. We don't have that problem because there is no bolus tracking you don't have to worry about choosing the bolus right through the coronary we can pick it up one hundred percent of the time with no problems even rabbits considering of course of the rabbit coronary artery is a lot smaller than the human coronary are. [00:20:01] Here's the thick slab rendering I was telling you about. So this is a two point five two point six millimeter thick slab and we can see the septum different chambers and we can start seeing the valve plan here. Equivalently displayed here also. This is again three and a half hours after injection of the contrast agent. [00:20:25] So essentially we're actors with our contrast agent you don't need any timing and you don't have renal toxicity I didn't talk about that a lot here but it turns out clear via the liver in the feces and so you can use this even if toxicity is a problem for a given patient because you don't send the agent to the kidneys at all the next thing that we started looking at was what can we do beyond the blood pool can we imaged tumors and it turns out that there is a wealth of literature on the localization of life zones to tumors stave been around for a long time. [00:21:13] And for the pharmacokinetics for the transport of life a zone from the blood compartment into intra to move to Merle capillaries and for the localization of the lie presumes in the inner city volume These are very very well known numbers and you can clearly see the clearance numbers are also very well known and you can clearly see the because this is one directional you're going to get a continuous accumulation of lipids arms into the interstitial volume of the tumor and therefore one would expect that the tumors would light up very nicely over the blood pool. [00:21:54] And that's exactly what we get when we simulate this model. So here's. The blood levels in Huntsville units again. And here's the prediction for what happens in a tumor that is undergoing angiogenesis which means extrapolation off the I presume is going to be a possibility. And so two or three hours post injection you can clearly see that the tumour is going to be lit up substantially over the blood pool and this has huge implications for early detection because what this means is any tumor where angiogenesis is taking place is going to be targeted will by a lie presume a contrast agent. [00:22:40] How soon do tumors start undergoing angiogenesis when they're starting to grow Well we started measuring that there's also literature date on it. So this is a slice through one of our tumor models. So here's a see six glioma in and rack brain and this is nine days after injection that's grown quite nicely but as early as two days post injection this is a factor eight stain you can start seeing factory. [00:23:13] Present on the inside of the blood vessels. This is for days. This is nine days. And so we're very confident that very soon after angiogenesis begins and that might happen even two days after implantation of these tumors for example that we'd be able to pick them up so the onus now is not on the contrast but the onus is on the resolution of the C.T. scanner and so what we feel is that we can pick up tumours at the size resolution of the C.T. scanner no problem. [00:23:49] Now what you're going to do with that once you detect it is another story altogether. Because it's not even clear today that even the traditional mammography screening to. As for example that Tara P. is the right thing to do. There's a paradox in breast cancer screening particularly for women between thirty and fifty for example. [00:24:15] Where if they were screened their mortality rate was actually higher than if they weren't screened and that's because adjuvant chemotherapy was not being used. They simply went after the primaries with surgery and then the secondary metastases started blooming. So what it comes back to is if we can detect these things at one one point two millimeter size. [00:24:38] It's not clear to me that Tara P. is ready for that yet but we can certainly detect them. OK so I'm going to change gears a little and stop talking about contrast for a while and start talking more about targeting. Because it seems to me that that is the next generation of cancer therapies. [00:24:58] We started looking at Cleo must fairly early on when we got into this because for a number of reasons. Apart from the high incidence it turns out that there are unique. They're deer diffuse and they're invasive. But they're also separated from the bloodstream by the blood brain barrier the tightly packed cell or lining it in to kill you. [00:25:22] This makes treatment of gliomas quite difficult to do and the most effective drug today B.C. and new is not particularly effective from a cell kill point of view there are plenty of other came with therapeutics that are much more effective from a cell kill perspective B.C.N. you works because it happens to have high blood brain barrier permit ability. [00:25:47] So the approach that we were taking was how can we enhance blood brain barrier permeability for some other agent that potentially has a higher self kill rate. As it turned out. We were interesting. We were interested in targeting otherwise for a number of other reasons. And we were looking at ligand receptor based targeting schemes and the thought that we had was. [00:26:11] Let's take our target of a kill and and shall we say transparent to the outside of the space. And enhance the blood brain barrier permeable A-T. duckling. And then of course we started thinking about the same things for contrast agents. So I call this a rational approach to Ligon targeting and again we started out with a life a zone basis for our carriers. [00:26:41] And it's been known for quite a while that if you put drugs inside lipids omes and then in order to maintain these things in circulation you put polyethylene glycol on there outside or whatever other hydra Felicity enhancing agent. You cannot put the targeting Ligon directly on the surface of the lipids own because it gets clouded by the peg you really need to put this targeting leg and at the distal end of the pegs so that they can match up with the receptors on the cell surface. [00:27:14] And so our approach has been to mathematically model the binding and the optic process and the idea here is that we can optimize the lengths of these peg heter So if you go back to the previous slide a question to ask is How long should the step that are at the end of which the leg and is bound. [00:27:37] How long does it have to be with respect to the rest of the legs. Clearly this is an extreme case where that length is zero and we know that's not effective is it effective if these testers are the same length as the surrounding pegs do they have to be longer. [00:27:52] How much longer are those for some of the questions that we were trying to answer. And also try to answer how many of these lagoons two we really need in order to. Have optimal binding and so we've developed actually a pretty good mathematical model I think of this process and were at the stage where we're able to fabricate carriers now with different data lengths and different bigger numbers we're testing these things in vitro and we're just of the point where we can start putting these things in vivo we're running into some practical difficulties there but it's not it's only a matter of time this been actually a lot of literature on the modeling of surface and cell interactions or surface and carrier interactions Berger's group has looked this is a relatively old work going back to early ninety's the binding of multivalent lie against to sell surfaces. [00:28:49] Linda Griffiths group has also looked at simulation of cell surface. But again the fundamental difference between these and what we're doing with targeting drug carriers. Is They're all looking at cell surface and he and so they're looking at relatively large cells and hearing to even more extensive surfaces. [00:29:16] There are a few pieces of work where they're looking at virus attachments to cells and one article that I have found actually on little zome cell interactions but the big shark coming with this is there is only a single interaction considered between live in the cell which means one leg and bonding to one receptor whereas And they're also look at conventional I presume so there's no quality in glycol coating on their outside so it's really not very realistic for an in vivo situation. [00:29:51] So the the model that we're building is really a political two Stealth low. I presume ST TNG humors we are going to is that receptors on the surface of to sell into clusters and you've got to take those dimensions into account or turns out. We'll talk about that a little bit later on we'll look at a lie. [00:30:15] Presumes that have polyethylene glycol for stealth properties and Liggins bound to the distill and those tethers just to position you inside space with all these different species life zones that we work with they're typically one hundred hundred twenty five nanometers or so in diameter polyethylene glycol coating on the outside of these live presumes is a few nanometers so that the flurry radius for a two thousand molecule or a polyethylene glycol molecule is on the order of three and a half nanometers cells are fifteen microns so they're huge compared to the life prisms so for all practical purposes a cell surface can be considered flat and the cluster sizes they vary quite a bit but a typical size is about seventeen animators So if you think about it a lie presumes sitting on top of a cluster pretty much saturated such cluster. [00:31:16] It's very unlikely that a different life a zone will ever be able to buy into the same plaster the number of receptors for a cluster there is a little bit of literature on that and I've seen numbers varying between fifteen or about forty. OK So when the life zone or drug carrier starts approaching the cell surface what happens like ends at the end of these tether start reaching out and grabbing on to receptors on the surface. [00:31:46] The first likelihood of such an event occurring is if this polymer tempter reaches its fully extended length and I'll show you a little later on that this is fully extended. Length is actually quite well defined for different polymers and once that happens it pulls the light presume in closer to the surface the polymers now a little more relaxed. [00:32:12] But still at this binding distance. There's a limiting number of like Ganz on the surface of the light presume that can really bind. So you fabricate these lipids ohms of leggins all over. Not all of them are going to be active and binding to the surface is only going to be a small fraction of them to her active for binding to the surface. [00:32:35] So here is a molecule or Dynamics' type calculation from Vladimir to our children's group showing the probability of existence at different lengths for polymers and this is reduced parameter that is proportional to the molecular weight of the polymer and the idea being that the larger this parameter against the longer the polymer molecule or weight and the larger the fully extended length noticed that these things have a relatively flat probability of existence until you have essentially this cut off dimension and it drops off like a rock and that's the fully extended length that we consider in the calculations. [00:33:17] When you put that in what you determine is that there are two crucial parameters one is the active area of the light was on. So this is the fractional surface area of the entire LIFE WAS own that is actually active for binding to the cell surface. And there is this radius of the area of influence on the cell surface so for example this little receptor is outside that radius of the area of influence and cannot bind to the Sleipner so it's only these that are within the area of influence that can bind to the life itself. [00:33:54] So you can calculate using relatively simple geometry considerations. The active area of the life a zone that have plotted here as a percent of the total surface area of the life zone versus the molecule or weight of the tempter and what you can see here is that the larger the lie presume gets the smaller the percentage is so for a twenty five nanometer life presume you could potentially have as much as ninety percent of the light was an area actually being active. [00:34:28] It's not a very practical situation because at twenty five animators there isn't a lot of drug you can get into the like zone. So let's take a more realistic situation around one hundred twenty nanometers or so but you can still get it if you use very long polymer chains you can get as much as fifty percent of the live bizarre area actually being active more likely you're going to be don't understand to twenty percent range. [00:34:55] Similarly you can calculate the area of influence on the surface of the lot of the cell as a function of Peg molecular weight for different life Azzam diameters and it has this nice straight line behavior. The reason I put that up there is I wanted to compare those dimensions on the cell surface with the sizes of features on the cell surface so cavalierly and clatter and coated pits and the lip of draft micro domains which are very common in the case of receptors that are fast for Title I know so tall and heard these all have dimensions and is fifty to one hundred maybe one hundred fifty or so and then a meter size range which compares quite nicely with the radius of the area of influence for polymers that are between Chile say two thousand and five thousand or so and Monica the rate which are the most common ones we're going to use. [00:35:55] So the this is the basis of the calculation that's as. For most a far carriers one life to zone carrier is going to fit on one hit and it's not really going to reach over and grab either different cabriolet sort of different and of course when you get to a very long molecular weights. [00:36:20] You can start reaching over from one pit to the next hip. And have binding with multiple pits and the question that we need to ask though when that happens is binding the for drug carrier to the cell surface play itself is not sufficient the drug needs to be internalized into the cell before it can be effective. [00:36:44] So if you take the case of a chemotherapeutic drug like dogs the rubes and positioning the drug on the outside of the cell by itself is not very useful. It needs to get into the cell right. Now if a life a zone is anchored to two separate plus structures on the surface of the cell and both these clusters are essentially going to try and internalize whatever is anchored to them what happens to the lipids omits anchored to two different clusters is probably going to get torn apart or in some fashion. [00:37:13] One of the Tattersall's going to get pulled out of the lip and by layer and it's not very useful. So there is a far chewed as optimisation that has gone on in picking lipids omes that are in this size range so they can anchor to one internalizing mechanism and be dragged into the cell as a result. [00:37:33] So the status of the current model is we have stealth like presumes would like Gantz we've taken a life size into account if they can get their length into account. We've taken receptor distribution on the cell surface into account and it turns out. We've also accounted for saturation in the uptake process and I'll talk about that a little later on. [00:37:55] But at a very high level for example we work with. Based targeting So we've got a lot of presumes that have full it's that they're here. They buy into the cell surface they get internalize pretty soon the cell says I've had in a full eight I don't need anymore and internalization rate based on full intercept or binding drops and you need to take that into account in order to be able to fit the data that you see. [00:38:23] OK. I want to talk a little bit about the different classes of receptors that we can target and I've classified this in a couple different ways one based on the internalization pathway and the other based on and distribution. So the full interceptor for example is one hundred percent clustered there are no free fall interceptors on any of the cell types that we have looked at. [00:38:48] They're all in clusters a number of Perceptor is however are not so cooperative. They have clusters as well as a random background distribution receptors and so if you're targeting one of these it becomes quite important to take and direct the fact that the cluster is likely what is going to internalize the free to scepter and needs to be recruited into a cluster first before it can participate in the internalization pathway the internalization pathways by themselves depend on how the receptors are located on the surface in caviar lay our anchor to G.P.I. platter and coated pits and this classifications available for all of these again going back to the internalization of whatever you anchor if you're using foliate you'd really like to pick like ants that all fall into the same mechanism. [00:39:48] Otherwise you're likely to end up binding to different features on the cells surface. OK so in order to build this model we first started modeling our experimental. Mental system which was done in tissue culture in Wells. So here's a cell monolayer on the bottom surface of the well there is a lip Azzam suspension that is in contact with it so mathematically that produces to simply diffusion of lipids ohms from the bulk to the media to vicinity next to the cell surface and we we account for that by a simple diffusion constant eye turns out we can do that and we don't have to worry about all the depletion effects here because the system does not get depleted number of life zones that we put in this so large that there's only less than a percent that are actually bound to the cells of it any time and so we can lump all the diffusion down into one constant which gets multiplied by a concentration term these lipids ohms are then going to react with clusters of Perceptor is on the surface with a forward and reverse rate to form a lip a zone that is attached to the surface through one bond on the surface of the cell ductile I presume that is bound with one Bond can then form a second bond and form a life goes on. [00:41:17] That is attached through to bonds and obviously this process can go on. It's a population balance type model and it has the same rate constants forward and reverse as the initial binding the lipids own that is bound to the surface can also internalize into the cell and so that's another flux pathway that results from the bound life presumes So here's both of them together. [00:41:45] Ultimately what this leads to is a reaction network that looks like this you've got binding to find a N. bond life a zone on the cell surface which can be internalized art can farm and. Plus one Bond life goes on and so you can think of this as a population balance model with depletion which is a fairly classical set off mathematical models for this for this type of system. [00:42:15] What that reduces to in terms of equations says a number of ordinary differential equations and I'm not going to go through every term in this message for you. This is simply the family of the population balanced so that each X. one X. through X. and are simply the single bone then the multiple The bone dry presumes on the cell surface and then the other species that are not necessarily bound are referred to by bulk concentrations and here are the O.T.S. governing those you can solve all these together very conveniently using any initial value problem solver. [00:42:53] We started out doing this in matlab at one point we use our own Fortran course for this now but at the end of the day. Once you solve that you solve that as a kinetic model and then you find the levels here is data from Phil Lo's group who was at Purdue at the time and this is Lippo zone binding to K.B. cells at four degrees centigrade and what happens at four degrees centigrade is that the internalization rate is is dropped virtually to zero the cells are not internalize anything at four degrees this simply bind and we can match the the binding results almost exactly provided we put in the diffusion constant in the absence of the diffusion constant we get a very rapid saturation and it doesn't look anything like the data at all. [00:43:48] Incidentally if you calculate out using the diffusion go efficient for the size ly presumes and the surface area. You end up with an overall rate constant at. Almost exactly this number so we were very happy with that value. We had to set the internalisation externalisation rates to zero. [00:44:07] Of course because there was no internalization So the obvious take home lesson from this by the way is for those a few who do micro tire plate or cell culture type experiments and uptake you have to agitate those systems you can't just leave them static because if you do your diffusion limited and what you're likely seeing is not a trend based on binding but more likely a trend based on mass transfer limitations. [00:44:39] So our experiments we actually agitate the plates and we have a much higher diffusion of a constant than this but anyway as we can fit this is exactly follows group also did experiments at thirty seven degrees where our internalise as well as bone and we can match that data perfectly well to it turns out that in this case the diffusion rate a slightly different because at a higher temperature the diffusion coefficient for the lie presumes is higher. [00:45:14] There's also a component for regulation of the fully Cassatt uptake but that's not so clear in this case I'll show it to you in data where the down regulation is much more obvious. OK for our own experiments we've constructed of conjugates that we can post insert into life as ohms not show you exactly how that is done and remote loading of toxin Robeson into the life of them so much time to have were running out. [00:45:47] OK so I'll go relatively quickly here. OK so we make our conjugates by actually a relatively simple car but I admit chemistry linking. And mean to the folic acid and farm the species. An issue that we had to deal with was if we were looking at multiple I guess our life was on the surface and of for example we were looking at transfer and it turns out that they are very different in size. [00:46:19] Here is a carton of folate molecule with respect to the transfer and molecule they are very very different in size and in order to present both these lie against to the cell surface at exactly the same place. We need to put these things on relatively sharp Tattersall and the foal it's on very long counters. [00:46:35] And so one of the things that we've recently worked out is a a standard chemistry to be able to do this and we do this using peg thirty three fifty building blocks and so this is a I guess a home a block or polymer approach that we use and so we start with the D.P.P. even the two thousand molecule or wait peg and then I draw sets an image of that and we can add any number of these thirty three fifty four late I mean groups to that and so the reaction of these two together is going to give you a D.P.P. with a peg fifty three fifty which is the addition of these two plus the full eight. [00:47:12] And so we can continue this any number of times in order to build a longer and longer chain at the end of the of the D.P.P. we also pull up a neat trick in putting the lie against into our life was ohms we pre-formed lipids ohms and then we add my cells off the leg and conjugated a lid and they simply insert into the life a zone. [00:47:39] The advantage of doing this is that we can very precisely control the number of flagons on the outside of the life zone. We load the drug into the lipids omes using the remote loading technique with sickly we put on one yourself it on the inside of the lipids ohm initially. [00:47:58] A dissociates out of. Free Omonia which leaves the life a zone because it is a neutral molecule it can diffuse out and you exchange the docs troops in which is on the outside of the life a zone in and it precipitates out as the sulfate and this is an extremely efficient process and we can load one hundred percent of the docs are obese and this way cell culture details OK. [00:48:23] So we did this on three different cell lines K.B. cells which are positive control I guess they overexpressed folate C six glioma added the nine particle cell line which expresses zero folded receptors and the reason we did this was we really wanted to see if we could differentiate between the C six and the nine cells because that's pretty close to the situation that you would be looking at in that where you'd have tumor cells right next to an other in your own cells and clearly we can differentiate them in terms of docs uptake quite nicely. [00:49:01] Of course the positive controls do a lot better. We've modeled this mathematically and we have modeled this downtrend also quite nicely mathematically here a number of other controls that we did in order to make sure that the binding was actually fully mediated So we were able to inhibit the binding by adding free folic acid and that is the other one here. [00:49:27] OK So and we also distant Co culture in rather than have these and separate wells we had both cell types in the same well and so we were incubating in the same medium and again we can see the same enhanced meant for the K.B. cell line. OK so if you go back to the K.V. cell uptake data. [00:49:46] It turns out that in order to have the same number of cells exposed to the medium. Just in the did these experiments needed to use different concentrations of. So in order to fit this data I really need to put in two different concentrations of life a zone is five then a mauler and seven then a motor and those are these two lines and once you introduce the down regulation of folate uptake we can also predict this reduction in the uptake as a function of the number of flagons per life. [00:50:22] So what the cell is doing is it's internalizing based on the number of fly presumes that are bound. And when it gets too much phone late at saying OK I've had enough I'm not going to internalize nearly as many anymore which is why you get reduce the number of like those ohms bond notice however that because this is a large number of folate per lipids own even though fewer light presumes where internalised are more full it's have been internalized that's why the cell has been saturated. [00:50:51] So if you plotted this in terms of full it's internalized this would be a flak. You can't. So also notice that our diffusion rate is substantially higher it's an order of magnitude more than an order of magnitude higher than it was in Phil Lo's experiments that's because we retain the agitated players. [00:51:14] OK so what I always ask my students this what's the big deal. You've done a lot of to swerve and then what we can predict the binding of life is ohms and the uptake of contents into cells. We can optimize targeting parameters so we can program into our model. [00:51:30] How long the polymer testers need to be for a given leg and size is a relatively general model that only looks a binding in internalisation so it doesn't it is not specific to two cancer therapy. We can use the same things for binding to various other sites inflammation sites and plaque so on the models also suitable and the platform technology that we have developed is also suitable for both Terra puting delivery. [00:52:00] And contrast to living in the future we're going to look at overlapping cluster stuff does tend to happen that's not accounted for in our model we have to put in to the lengths of the non-league untethered pegs. How do they affected area influence because they tend to make the did the length distribution of the lagoon tattered the tether Ligon pegs longer. [00:52:33] We have to look at like invalid C. of course and the next generation really is what Robbie Belle I'm going to and I call the multiple like and system and think here's a cartoon of that. So the idea that we're going after here is we've had this longstanding interest in the recovery mission of surfaces and how do you recognise it given how the cells recognise surfaces in the body for cancer therapy people have always looked for this one unique receptor and said if you can target that receptor then you can get your drug to the tumor. [00:53:08] Unfortunately that one unique receptor has been somewhat elusive You can find unique markers that are interest cellular but it's very rare to find a marker that is expressed on the outside of the cell and E R B two is one of the few that are that are reasonably unique for it to breast cancer for example. [00:53:30] So the approach that we've taken is to look for more than one like and let's not restrict ourselves to one leg and targeting wondred just scepter but maybe the cell surface has some unique fingerprint that is a combination of receptors. So can we tailor our drug carrier with many different species of ally against on the outside which would match up with the fingerprint of the receptor profile on the cell. [00:54:00] Surface and we've actually modeled that although I haven't shown you the data today and showing the good tool I guess we get roughly a thousand for them. Hands meant and binding over a single like an system and so the next step. Obviously is to fabricate those things and try out a multiple like an experiment. [00:54:21] OK I'll take questions and I'll leave you with a three D. volume movie of four heart. Robert heart that we got using our long circulating contraception. Thanks Mr stunned one receptor draft time. Great way for true to hundreds of types of cells can we just not included go to separate mobility interceptor recruitment into the calculation has a few know G.P.I. a bone grosser person are likely to be fairly mobile. [00:55:17] It's not clear what the signalling pathways are of that cause the recruitment at least they're not quantitatively understood that some of the reactions may be understood but the rate constants from the signalling are definitely not understood as of now and that's the primary reason why we have not included in the calculations mechanistically it's actually quite simple to put into the calculation and as soon as we know what those reactions are we will plan to blend them in our lives. [00:56:02] I wonder what there is to monitor monitor using now. Any time you want to any time you want to look at processes to take place over that time is the obvious answer but that does not have to be a biological process it could be something as mundane as workflow. [00:56:30] In a radiology clinic. It might be easier for them to simply injected one time and get the Stanley Iran. But for example the chill murder to action example to show to them. You see in hands meant of two more contrast over the blood at three to four hours after injection. [00:56:54] So any time you want to see things leaving the blood pool of them going elsewhere. That's when you would really want that long circulation time all the other you know of course it's depended on the identity of the of to individual like against and so plugged in on dates and offer it's for a particular leg ends in the calculation a particular calculation we used relatively naive useful it and transferred and it turns out that's probably not a good model to use because there are different classes of deceptive as an internal ization is not likely to be ID or fish and. [00:57:47] And so of the I would have to revise that unfortunately those binding constants are not easy to come by either. So we pulled them at rare intervals from him but the short answer is very dependent on the leg and receptacle. All right. Thanks.