 In last lecture, Dr. George LeBair gave you an overview of proteomics field and to perform high throughput proteomics based experiment, they need to generate the clone repositories and to achieve the high throughput biology and performing proteomics experiment, what are the key considerations you need to pay attention for generating those clone repositories. Once you have obtained large number of gene clones, now you are ready to perform many experiments and one such experiment was developed in his lab which is a novel protein microarray technology which is NAPA or nucleic acid programmable protein arrays. To perform NAPA, if you have these clone repositories available, you have large number of cDNA clones available, you can print them on the chip and imagine that what goes on in our body to do the central dogma, the transcription and translation process from the genes to RNA and the proteins. The same cascade of event could we try to reproduce, could we try to replicate on the chip itself and that was a concept of NAPA. From the these cDNA sequences printed on the glass slides, could we add the machinery which is required for in vitro transcription and translation and use those to synthesize the protein on the chip itself looks science fiction, but it was a reality. Dr. Josh Labair and one of his senior postdocs Dr. Neeroshan Ramachandran, they made this technology for performing high throughput protein microarrays using NAPA. Today Dr. Labair will talk to you about the development of this NAPA technology and how to use this for various microarray based applications. So, let us welcome Dr. Labair for his lecture on NAPA technologies. All right, so we are going to start now by talking a little bit about the NAPA method right and I already I already spent some time talking to you about the gene cloning part right so that is how do you make the clones for the genes that you are going to put on your protein arrays. We will talk a little bit today about NAPA production and I am going to do it a little bit from a historical perspective so how did we come about this method and how does it generally work then we will talk a little bit about how to do discovery on the platform and then finally nothing you do works if you do not go back and validate it. So, that is kind of like the whole end to end process so let us talk a little bit about protein microarrays right what are they and why do we want to do them and I am going to point out that these are a number of the people that were most responsible for doing this work Neeroshan Ramachandran was really the leader in our group that really pioneered this methodology Jeanne Haynesworth was the engineer who did a lot of the work and then some of these folks also contributed to some of the early methodology. So we talked earlier about how having a library of expression clones would allow you to do all kinds of different studies well here is one that we got very interested in which is biochemistry. If you want to study the biochemistry of a protein you need to be able to make that protein and do experiments and if you want to be able to do it times thousands then that is what led us to the idea of protein microarrays and just to give you some perspective when we first began this work we actually started by trying to do high throughput protein purification and so we were developing methods to make proteins in bacteria, life bacteria, capture the proteins on columns, elute the proteins and then study them in high throughput and believe it or not we still do some of that work for other reasons but we realized very quickly that it is hard to do high throughput protein purification that if you try to isolate lots and lots of proteins first of all a lot of the proteins don't purify well secondly the ones most of the proteins don't give you very high yield and then third you don't really know if the proteins that you're purifying are going to be of high quality folded you know an active. So that's what led us to this idea of protein microarrays and there are two kinds of protein arrays the first kind which I'm not going to spend a lot of time on are these antibody arrays where you print an array on a microscope slide you put antibodies down that recognize different proteins on them and then you use those arrays to probe a sample to capture whatever proteins are in that sample as a way of measuring the levels of those proteins in the sample. So the goal of this array is to measure the levels of proteins. The protein arrays that I'm going to talk about are called target protein arrays and the goal for these protein these arrays are to look at the proteins themselves what do they do who do they interact with how do they fold what is their function. So the idea on these slides is that you have a slide and each of these different spots represents a different protein on the array. So lots of things that you can do with a target protein array you can look at drug specificity I'll show you an example of that later you can do biomarker discovery you can do enzyme substrate identification you can do interaction domain mapping you can do analysis of how protein mutate of gene mutations affect the function of the proteins you can look at off target protein interactions. So here's an example of what that might look like imagine if you had a fluorescently tagged molecule and you probe the array you can see which protein that molecule targets or if it turns out that it binds to multiple proteins you might get it binding to multiple proteins and you might see the differences in the binding and that would give you some sense of its specificity. This is the area that that I've spent a lot of my career on and that is looking for patterns of binding that indicate the presence of disease I told you yesterday that we were looking to do look for markers for breast cancer right and we did that by looking at the immune response those markers. So imagine you take this serum and you apply it against the affected individuals or the normal individuals now I've shown that a number of spots light up and I did that for a reason it turns out that even normal people develop antibodies against some proteins the problem is that the word normal is in quotes it's in quotes because all of us have medical histories we may not have cancer but we have had other things in our lifetime and those things can affect your immune system and those things can give responses. So the key here is to know which responses correlate to cancer and which responses are not related to cancer right and so you have to accept the fact that there will be these other responses and then the idea is you do a variety of informatics processing to take these compare these to these and look for patterns like this one here that's present in everybody look for these guys that occur only in the affected individuals and then there will be some that are just random variation that occur from person to person. So that's so that's another approach that we're doing so so let's start here by asking the question what are the ideal qualities that we want in a protein array what would make an array a really good array so I would argue the first thing has to be high density the whole point of a protein array is to get lots of different proteins in a very small space so you can study it of course you want to be able to work with small volumes the advantage of a protein array is that you can take only a couple hundred microliters or a few microliters of serum and test thousands of proteins with that informate with that amount of course it needs to be multifunctional you'd be up to be able to test it with lots of activities but then you you also want it to have natural folding you want the proteins on the array to look like they do in normal circumstances and ideally they would be made in a milieu similar to the one in which they normally occur. You don't want to have to purify the proteins because if you have to purify the proteins then you're going to end up with all the things that can occur during protein purification proteins could lose their folding they could get low yield they may not be in the proper confirmation. You obviously want to be able to test as many different repositories as possible so you want to be able to test any proteins that you'd like and you'd like the the to not worry about the shelf life of the protein. Once you print the protein on a chip there's always this worry that the clock is starting and the longer it sits there the more likely it is to stop being active or being well-folded. All right so and of course you want the levels of protein from one to the other to be very consistent those are the things that you'd really like. All right so so this is the array type that we're going to talk about today which is called NAPPA for nucleic acid programmable protein array and the idea for NAPPA is that we print the gene for the protein on the chip and we store the chip as a DNA chip so the gene is there the clone the ones that we talked about in the first half of today is on the chip along with an agent that's going to capture the protein when it's made but the protein hasn't been made yet then on the day of the experiment we add a self-reextract make the protein and capture it and then we we display the protein at that time fresh and so the idea of NAPPA is that we can do all kinds of studies on it we can do interactions with specific protein queries we can do enzyme substrate modification we can even build multi multi protein complexes so let me begin at the beginning by showing you the the idea behind NAPPA before it was actually on a protein array so here what we're looking at is just making proteins using self-reextract in the wells of a 96 well dish okay and so the proteins that we make all have a GST tag at the C-terminus remember we talked about you always have to have a tag so these guys all have a tag and so if we if we make the protein in these wells and we probe them with an antibody that recognizes the tag they all light up right but if we probe them with antibodies that are specific to each individual protein then only the p21 lights up at p21 only the 16 lights up at 16 only the FOS and so on and so forth so depending on what you use as your antibody you only get that signal now you can use the same approach to look at at interaction queries in this case here's a bunch of different proteins on a chip right and in chip I mean in quotes because these are actually the wells of a 96 well dish if I probe with a protein p21 it binds to all these CDKs it binds to these cyclins it binds to these guys here and it binds actually to itself again so those are the interactors for p21 if you probe with a different query in this case CDK4 it's going to bind to the cyclin D's and it's going to bind to p16 and if I probe with p16 it's going to it's going to pick up these CDKs so you can see using different queries I get different interactors and again this is not a microscopic array this is a 96 well dish well of course the advantage of this is now I don't have to express these proteins or purify them they were made using mammalian extracts the levels of proteins were consistent on the on this chip they were made at the time of the experiment so I made the proteins and I tested them minutes later and of course I could do this general approach using any kind of CDNA if I can clone the gene and make the CDNA I can make the protein array and then we talked about the multifunctionality and of course users can modify this as they need to but there's still some challenges so we need to make we need to now take it from this format which is on 96 well dishes and we had to get it onto a microscopic chip we had to be able to print it on slides so that we could do thousands at a time so we had to be able to deal with very low amounts of protein we need to make at least 10 femtograms of protein we had to get capture that was rapid we had to find arrays that were compatible with standard array readers at the time we really wanted to work on a single slide format that that what that where we could add the extract to the whole slide we didn't want to have to manufacture these specialized methods for expressing in little tiny wells or something like that I'm going to come back to that because we now are moving in that direction but at the time we really wanted this to be very simple and of course we had to avoid crosstalk from from from spot to spot so let me talk about the technology so most people who built protein arrays do it by purifying the proteins first they they do what we started with which was this high throughput purification kind of technology and they do it in 96 well plates or even 384 well plates but they have a couple of problems first of all you get very highly variable protein yields so the amount of protein that you get from from one protein to the next can change over four logarithms so four orders of magnitude of difference they get they're working typically in heterologous systems so they're either purifying in bacteria or they're purifying from yeast or they're purifying from insect cells and that in and of itself introduces some differences and of course the biggest concern I have is a race shelf life so you purify the protein you store the protein then you take the protein out then you print the protein and then you store the printed protein so you have a lot of steps in there and all those are opportunities for proteins to lose their shape lose their folding and not be as functional and of course some proteins will stay fine during all that process and other proteins will not and you never know which ones they are there's no way to tell which ones are the good ones and the bad ones so this is the idea behind Napa in the case of Napa what we do is we print the gene on a plasmid we add cell-free extract that makes the protein and this is meant to show that we have the protein in blue and the GST tag in red and then here's a different protein in yellow and a GST tag in red so we make this at the time of the experiment and then what's gonna happen is yeah I just told you that is that the GST is gonna get captured by the anti GST antibody and now you're displaying the protein on the surface so it flips upside down so that the protein part is what's facing up make sense and then this is what it looks like here's an early array printed eight different protein 64 times each and here we probe with anti GST that's that's a way of measuring how much total protein we have on the array and so all of them light up and then if we probe it with anti p21 just remember the little 96 well plate same idea now we only get the p21 lighting up all right and I already kind of covered these inner these these advantages pretty much the same ones here so I won't go over them again all right so so here's how we first tested this we decided to do protein-protein interactions so imagine that you have three different spots on your array in fact we had we had many more than that but let's just talk about three this one makes the yellow protein the red protein and the blue protein and these are the genes these are DNA and then here's the antibody that's going to capture the GST now we want to we want to ask do any of these three proteins interact with the pink protein all right and the and the case in this case what we're doing is we're going to add the gene for the pink protein in the solution along with these guys which are attached to the surface so these guys are bound to the spot this is free to go anywhere it wants we then express the protein right and these three proteins are going to get captured to the surface of the array because they have the GST tag that's going to lock them down to their spot but this does not have a GST tag so it's going to float everywhere on this with across the array and then over time if you give it time the query protein will bind to the target proteins if it recognizes them in this case it binds to this guy but not these guys right and so now I can wash away anything that's not bound and I'm left now with this guy bound to this spot now how do I know where it binds well I know the identity of every spot on the array I know whatever position it is which gene it is and so I know that if this spot lights up that pink binds to red right and I can detect that interaction with an antibody that has a fluorescent marker on it that will recognize either a tag on the pink protein or it could recognize the pink protein itself and there's all kinds of variations you can do here you could use click chemistry to look at interactions and we've done that you can have you know other molecules that interact with this guy avidin and biotin lots of different strategies but the bottom line is as long as you can recognize the query protein you can determine where the binding occurred so our first experiment was this guy we took all the proteins in the human DNA replication complex these are this is that collection there cloned all those genes and then we printed them and expressed them and this is measuring with GST just to show that they all got made okay and everybody was done in duplicate so they were all there in two spots and then we we would query the array using an antibody using this protein here one of the proteins in the set MCM 2 and and then you can see that MCM 2 binds to work 5 or or 6 and MCM 3 and here it is binding to these guys over here and of course it does it in duplicate so we know it's we were confident of the result you can do the same thing with a different protein this is or 3 and again it's binding to certain proteins but not other proteins so you have every other protein on the array is sort of a negative control right and you can merge the two images if you wanted to and even build use that kind of thing to build an interaction map for all the proteins in the complex with all the other proteins and that's effectively what we did we identified we queried over a thousand possible interactions we identified 110 of them including many new ones using this general approach so you can use this to kind of look at protein protein interactions and of course you're not restricted to looking at full length proteins if you want to map the binding domain of specific parts of proteins you can do that so in this case we were looking at where does this protein Geminin bind to this protein CDT 1 so we took CDT 1 we made a series of different deletions right and we showed that all of them were expressed on the array and then we probe them with Geminin which interacts with them and you can see that Geminin binds to some of them but it doesn't bind to others so that gives you some sense of where the binding site is right this line here if that if this part of the protein was present then it always bound so that map quickly where two proteins may talk to each other and then Nero went back and made a very small version of this guy and showed that it was sufficient for binding the other thing that you can do if you want to play with these arrays is you can actually you can look at the possibility of multiple proteins interacting together so we knew that CDT 1 bound to the MCM complex we could tell that by looking because of biochemical studies that have been done before we got involved but we did not see CDT 1 directly interacting with any of the proteins over here this is a map that came from that big map I showed you and what we did what we figured out was that although CDT did not bind to any of these proteins it did bind to this protein and this protein bound to that protein so maybe this protein here is acting like a bridge protein it's holding CDT 1 in connection with that complex right so the question was could we test that on the arrays right and we did that by by doing a couple of things we we could probe MCM 2 we knew was in this complex so we've probed MCM 2 against CDT 1 either with CDC 6 or without it and we as a control had MCM 5 and we also had a negative control CDC 45 and this is just to show you what that looks like this just shows you that all the proteins were made on the array if we add MCM 2 without CDC 6 you don't see any binding here but you do see the positive control MCM 5 indicated if you add MCM 2 plus CDC 6 now now you can see the CDT 1 binding it's pretty faint I'm not sure you can see it where you are but we definitely observed it okay so we're going to spend more time later in the course talking about the high density arrays but I want to give you a flavor of what we had to do to get now from these arrays which I showed you showed had around you know 50 proteins on them the goal of course I told you from the beginning was to get to thousands so how do we adapt the platform to get to thousands so yeah so we were working in sort of dozens range we need we were we at those days we were using maxi prep DNA if you're going to do thousands of proteins and you remember that what we need to make is DNA not protein and that's advantageous because it's easier to make DNA than protein easier to purify it and much easier to quantify how much you've made but still as easy as it is to make DNA if you want to do an array of 10,000 proteins then you have to prepare 10,000 DNAs so you need to you you can't do that by maxi prep you have to be working at small scale that's reproducible and reliable yeah we were using this stripped avid in biotin chemistry to do our arrays at the beginning and that clearly was not going to work because it's expensive and it involved having to stick them in the UV box for a while it was really kind of a pain in the butt yeah the UV cross linking part was a cumbersome and we need to do some optimization and we needed to increase our content so I'm going to fast forward a little bit I'm not going to go through all of the optimization experiments that were done you're going to learn how to make these arrays using our current approach but we spent a lot of time working on the high density printing developing a better capture chemistry figuring out how to make DNA and high throughput and and figuring out how to detect the interactions in a in a more ready and easy way and this is sort of the result of all that work what you're looking at is kind of our current version of what we do most of the time for Napa now on the left is one of our typical arrays around 2300 genes or proteins printed on the array the signal you're looking at here is the DNA signal so we stain the chip every time we make one for DNA and that tells us that our printing was good right because remember what we're printing is DNA so if we stain for Pico with pico green and we see even staining that means that we did a good job of printing even amounts of DNA then we convert it to protein and we measure it with anti GST and that tells us how much protein we have on the array and that tells us that not only to be print well but we can express and capture the protein as well all right and that's what this plot shows you here in the x-axis is DNA signal that's the Pico green in the y-axis is GST signal that's the amount of protein made couple things that you can observe here first of all in terms of the x-axis the vast majority of our spots line up very evenly around this area here that means that we do a pretty good job of printing the same amount of DNA for every spot and that's encouraging means our printing is good you'll see that there's a few down here that did not print well and so it's good to know that secondly if you look at that from the top of the expression to the bottom of the expression the entire range is within one order of magnitude so instead of those protein arrays where you would have the lowest amount of protein to the highest amount of protein being over four orders of magnitude now all of our protein is in a single order of magnitude in fact 93% of these spots are within two-fold of the mean which means that we're getting very even levels of protein on the chip and that's exactly remember that one of the things we wanted from an ideal array was to have very consistent even levels of protein and just to give you some better sense of that here's quantification based on different types of proteins going from weakest to strongest and then this line here represents the low the lower end of detection and this is the higher end of detection I know that's not true here I'm sorry that it's right here I don't have that on this graph this is just to signal intensity but you can see that we get by and large 96% of transcription factors were detectable 97% of kinases membrane proteins are very detectable and then small medium and large proteins are all detectable so roughly speaking about 97% of whatever we print we can get good expression of occasionally we'll run into proteins that have unusual amino acid sequences that make it difficult to get high yield but that is by far the exception so this is this is the method that we use to purify the DNA or that I should say the method that we used to use those of you who know Sanjiva this is when he was in the lab this is how he did it this is an automated platform we had worked out using a robot how to do DNA mini preps using robots and that allowed us to do if you were really working hard about 600 a day in a sort of a team approach that that was not easy but you could do it just to give you some frame of reference when I did DNA mini preps in my day if you did 50 in a day you were working your butt off but with robots you could get up to around 600 okay since then in Arizona now we have this technology so we've taken that robot technology which was we used to call it sneaker net which is you connect one robot to the other by a graduate student who runs from one to the other to an a fully automated platform and let me see if I can make this go did that is that going yeah so this is what we have in Arizona now in the basement where my lab is located this is a fully automated automated platform for growing DNA and purifying it the camera is sitting on an incubator shaker that grows the bacteria with the DNA in it this particular shaker has a specialized door in the back so that these robot arms here can reach in and take out each one at a time as it needs to and they can put they pass it back and forth on that platform there so this guy is handing it off to this guy so he's gonna pick it up now but they can pass it back to the centrifuge which is right there there's also a freezer which is right over here that stores the pellets after they've been grown and then this liquid robot over here will purify the DNA from the bacterial pellets you can see this guy's mixing by turning it upside down just the way you would invert a plate right so the robots can do that for you if you want them to and of course we have sealers and peelers to seal the plate so they don't spill barcode reader there's a barcode reader right down here which you can't see that checks the barcode of every plate to make sure that it's what supposed to be and then that this this device over here will read the the optical density of the DNA after you've made the DNA the OD 260 so we can actually measure how much DNA we're making and we can the robots will automatically adjust the concentration to make them what you want them to be so all in all we went from doing 50 mini-preps a day to 600 mini-preps a day to now 4,600 mini-preps in 70 hours start to finish growing bacteria to getting purified DNA so it really accelerates what you do and also gives you a little bit more certainty that you were what you're doing is working so this is how we actually make the chips and I think you guys have examples of this at your desks you have this cover slip here that you put on the chip you there's a little hole here and a little hole here you inject your lysate in here you fill it all up make sure you don't get any bubbles that's always the trick one of the things you have to learn how to do and then you'll you can make the proteins on the chip using this approach this just indicates that you can map a more high-dent higher throughput version of mapping where binding occurs here's an example of an antibody that was binding to the p53 protein we did a series of n-terminal deletions and you can see it binds to all of them until it gets to here we did c-terminal deletions and again it binds to here then you don't see it and we did fragments that walk across the protein and of course it binds to just that one there so there's been a couple of of advances that that have occurred in the last I would say several years when we first started this work we were using reticula site lysate from better from rabbits to make proteins that was the expression lysate that we used to make proteins and I will say that still works okay this is what that looks like so you've been seeing arrays that look like this now most of the day blue is okay expression green is better expression orange is really good expression red is like amazing expression okay so that gives you some flavor of the thing and we were quite happy with this but then the patent ended on that and new companies came out and a new version of lysate came out that was made from human cells the advantage of the human lysate was that it it came from a purified cell line so you didn't have to get so much variation from animal to animal and this is what the signal look like just unbelievable signal about 15 times stronger than we could get from the rabbit lysate this is whole human extract that includes human ribosomes and human chaperone proteins which means that the there are proteins in the lysate that help these proteins fold in their natural in their natural shape the other advantage of the human lysate is that it's less likely to inhibit immune reactions so one of the problems that we used to see with rabbit retic lysate was every once in a while when we were looking at this is a person who was vaccinated with an antibody to these to this anthrax that even after vaccine we saw no response on the chips if we made the same chip with human lysate we saw very good responses probably what's happening is in the human lysate I mean in the rabbit lysate there are because it's from blood there may be inhibitors of immune response there and those were blocking it on the chips but the human lysate comes from a purified cell line there's no immune system around and so you don't get that problem so I'm going to show you just a couple of applications of this approach and then maybe even just one so we talked about immune responses the classic immune responses by Eliza where you you coat the well of a 96 well dish with your protein you add serum to the the the wells and then here's a patient who had a strong response and here's a patient who had no response of course if you do Eliza's you're doing one protein at a time it typically requires a lot of protein to do that and and and some proteins are very hard to make to begin with so of course what we'd like to do is this take a chip that displays thousands of proteins that okay so the idea is you probe a chip with serum and then various spots on the chip light up and we've talked about all these different types of assays so I won't belabor that so so let me give you one example we'll talk about more as the course goes on this is pathogen proteins right and what you're looking at here I remember I mentioned tularemia when you asked your question that is the entire proteome of Francisella tularensis in fact it's the entire proteome and duplicate we got them all into a single array so because we in the end we got that to work these five chips are cholera so Vibrio cholera and then this chip here is just outer membrane proteins from an organism called pseudomonas originosa which is the organism that causes pseudomonas pneumonia's it's a leading cause of death in patients who have a disease called cystic fibrosis I'm going to spend a little bit more time on this guy so there's around 300 proteins on there we were working with a collaborator when we were in Boston Steve Laurie was interested in identifying developing a vaccine against pseudomonas because this was the leading cause of death in patients with with CFN also a major cause of death of patients in the hospital who are intubated who are otherwise immune compromised this is an organism by the way that we've all been exposed to it's in the environment all the time most of us if we're healthy don't get infections but under certain circumstances you get infections so he his idea was the proteins on the outer membrane of the bacteria are the most likely to be inducing an immune response and and to respond to a vaccine so he wanted to look at which of those proteins was immunogenic his idea was he was going to purify those proteins and then test them now if you've ever tried to purify a membrane protein you know how difficult that is it's hard to purify proteins in general but purifying membrane proteins is a is a nightmare and so he he boldly went ahead to clone to purify 300 of these and I think he got four right so we suggested try the array because we knew the arrays can make membrane proteins pretty well and in fact they did so here's the chip the DNA stain here's the chip the the protein stain you notice that they're all red and they're all expressed so the membrane proteins were well made on the chip and then we probe them with patient serum and you can see that this patient is responding to certain protein on the chip so then you can ask the question well are there common responses because if you're going to make a vaccine you want to be one that's common that works for most people yeah so here he took a number of patients with cystic fibrosis who had documented pseudomonal infections here's a group of non CF patients who had who still had documented infections but just they didn't have CF and then here are healthy controls this is just to show you that the responses were very reliable from chip to chip and if you start looking carefully at this you'll start to see a pattern emerge certain features show up over and over again so that pair there is there it's there it's there it's there and there and there and there same it's those two spots are the same protein everybody's not there and duplicate they show up repeatedly that's a sign that that particular protein is what we call immunodominant and if you look at patients in columns and antigens in rows you'll see that these top 12 or so antigens show up in numerous patients and so this is a group of proteins that we should be looking at to think about developing a vaccine now there's a couple of things to remember first of all you can't make a vaccine so you know that those proteins actually induce a protective response so we haven't done that part yet the second thing is that no single protein worked for everybody it turned out that to get everybody you had to get a mixture of a few proteins and I think that's going to be a common in fact I know that's going to be a common theme moving forward in developing biomarkers is that it's going to be rare that a single biomarker will work for everybody eventually you're going to need multiple biomarkers because different people have different responses so I hope you got a very good overview of how this fascinating technology nucleic acid programmable protein array was developed the kind of thought process of generating these resources especially protein without having the protein expression and purification was differently one of the revolutionary concepts in the proteomics field the Napa chemistry was explained by Dr. LaBeurre in detail and you're also now familiar with what are the advantages of using this technology platform of course there are challenges in miniaturizing these assays these features to do the high density printing but those were overcome with many innovative ways and Dr. LaBeurre has talked to you about high density printing the new capture chemistries the different modified ways of DNA preparation and the improved detection technologies which have really progressed the initial versions of Napa technology to the very latest much more easy and reproducible and high support Napa based platforms I hope now you are very convinced that using self free expression microarrays could overcome many limitations of protein expression and purification you need not to limit yourself to express and purify each protein of interest and large number of proteins to be purified before you can do a protein micro experiment even if you have cDNA for the genes of interest you can still do the protein microarray based experiments and Napa could be one of the very powerful technologies to do these kind of experiments I hope you got some understanding of this novel technology and a basics of some applications which could be performed using Napa arrays thank you very much