 Hello students. In the previous lecture, I talked to you about using protein arrays, how one could screen autoantibodies from the brain tumor patients. Ideas to give you the workflow, what is involved of doing the actual biological experiments, actual clinical experiments where one could try to obtain some meaningful insight from these protein chips or different type of proteomic technologies. So, in the last lecture, I essentially talked to you about the workflow involved in doing these experiments. And then I talked to you about one case studies on gliomas, serum sample, how one could probe that on the chip and try to find the potential biomarkers for autoantibody screening. Let me continue a bit more on that because it is important for us to know that same platform could be used for variety of similar diseases and also one could not only use the serum sample but also other type of bio fluids. So, I am going to continue that application further and talk to you about another disease right now which is meningioma. Just to give you brief context to that, the meningiomas they derived from meninges of the brain which is you know the duramatter, arachnoid and parameter just go back little bit basics of your anatomy and think about the geology and look at you know the brain structure and you understand that you know while it looks like broadly they are all brain tumor but they within the different regions of the brain these kind of tumors can arise and actually their aggressiveness and the way of treatment are very different depending on the location as well. So, these tumors which are meningiomas these are most common, most predominant type almost 37% of among the all these CNS tumors. But other tumor which you can see the bottom panel which is the gliomas while they are all you know the not the highest numbers but they are the most aggressive like the glioblastomalty for me they are most aggressive tumor type and they have the origin differently which is from the gliocels. So, meningiomas in general they are slow growing well differentiated and demarcated they are mostly benign but of course as the disease progresses towards the grade 2 and 3 they might invade the brain and then you know one could have even the various malignant effect and potentially the death of the patient. But that is not so common as compared to if you think about the glioma type of tumors which are fibrillary and diffuse in nature and they are very aggressive and very malignant and they migrate and invade the entire normal brain tissues. So, I hope you got some idea that you know there are different type of tumors and when you are obtained the patient serum sample from different type of brain tumors there is a curiosity that is the same type of antibodies are going to be seen or are there going to be different set of antibodies which are going to be showing the signal in response to glioma differently and meningoma differently. So, that is the idea for this experiment where if you see on the screen we use the same workflow as I talked to you in the last lecture you know like a western blot scheme we have the protein arrays we probe the patient serum sample on the chip which is having the primary antibody and now if the antibody antigen reaction happens then we are going to add the secondary antibody which is anti human IgG linked with the Psi 3 or Psi 5 fluorophores and then you can edit those signals. In this case for the meningoma we had the patients from the control and the grade 1 and grade 2 and an idea was can we find out some distinct pattern of the proteins which is unique for the meningoma auto antibodies. So, very briefly when we get these samples from Pratay Memorial Hospital where our collaborators are there from the clinical side before the sample come to us all the you know the patients information is available where each patients immunohistochemistry data is available and based on those information they do the WHO grading. So, the patients belong to the grade 1, grade 2, grade 3 and also their MRI images are there. So, when it is looking at the radiological based features of the tumor as well. So, now in this case we can see that the tumor is in which location is that based on a skull base or supertentorial. So, while now we are going to screen the patient serum sample from the meningoma different grades, but we already have information whether the patients belong to the which grade of the tumor and what are the radiological location of those tumors. So, once we do the screening but as you can understand in the protein arrays we are taking one chip and one patient serum sample. So, we are generating the data which is all distinct. So, one could reanalyze the data in different manner based on the WHO grading scheme or based on the radiological base classification. So, after doing the screening we first wanted to ask the question based on the radiological location of the tumor can we identify the protein signature which could segregate patients into the radiological subgroups. And interestingly some of the proteins like vimentin, myosin 9, fibronectin and ribosome binding protein they showed the different differences even based on the location of the tumor which is a skull base or supertentorial. When we started looking at different grades of the tumor now based on the WHO criteria we found some very interesting significantly dysregulated autentic bodies in meningoma patient. And you know I have shown on the left side image one of the protein cryome which is highly expressing in the healthy individual but in the disease of the brain tumor patients of meningoma you can see no signal is seen. So, there is significant down regulation but other protein like MAPK3 or EFCAB2 proteins they are showing that in most of the meningoma patients the signal is very high very intense. So, of course, some of these are the interesting observation which one would like to follow up validate and look for are there you know the distinct signal from the meningoma patients which one could you start utilizing it for the reduction of these kind of diseases. On the right side what you see those are volcano plots where we are like after doing all the analysis we are looking at which are the proteins emerging from the protein microarray platform which are quite significantly distinct when we compare from the healthy control with the meningoma grades. Further what we are looking at you know how some of these protein which are significantly disintegrated or significantly altered from the healthy individual to the great patients can we also validate these proteins using western blot. So, for some protein which we have antibodies we try to validate those and especially if you see the selenium binding protein 1 and TPD 52 proteins for those we also try to validate their levels using western blots and we found quite similar trends in both array platform as well as the western blot images. Further looking at all the alternative bodies which are changing from the healthy versus meningomas and healthy versus meningoma grade 1 or healthy versus meningoma grade 2 we try to map all those proteins into different type of enrichment analysis and we looked at which are the major pathways which are perturbed. So, we found there are 48 pathways where most of these proteins are mapping which are disintegrated in meningoma grade 1 as compared to the healthy controls and some of these pathways were quite relevant which are shown in this slide here which are based on the map kinase cascade, EGFR signaling, osteopantine mediated events, signaling by NGF or signaling to RAS. Many proteins like NRAS, MAPK3, MAPK1, PTPN11 some of these are already implicated in majority of these pathways and also some of them are known from other independent studies that they are quite relevant proteins in the disease biology of meningomas. Further we try to enrich these various candidates looked at their interaction networks and found that meningoma grade 1 gave us 4 distinct clusters of protein interactions whereas the grade 2 only gave us one cluster for the looking at the protein interaction network analysis. These are shown on the left side and the top right side. Further we have also utilized the geo terms emerging from the healthy control versus meningoma grade 1 and try to map these interaction networks in this figure which you see in the bottom panel. I think it is important to appreciate the different technologies are giving us you know certain clues which are quite relevant to look for any unknown factors for that disease but it is good idea to also see how robust those signals are. Can we measure those proteins from different platforms and still see the same pattern. So we took the same patient samples from the serum of meningoma and then apply those on the mass spectrometry based platform and looked for how many of those auto antibody we can also detect when we look at the serum samples from the mass pack. And this table shows that there are many protein where we could now start comparing our signal from the tissue proteomics versus auto antibody response across different grades of you know the grade 1 and grade 2. And interestingly many protein which showed higher up regulation even from the tissue proteomics using mass spectrometer they were also seen higher from the serum proteomics of the auto antibody response using protein arrays. So these are protein which are differently quite interesting and needs further investigation and validation on the larger patient cohorts. Likewise this slide shows we also had some proteins which were down regulated and several of them also showed common pattern when we measured them from the mass spectrometers or from the protein array platforms. Finally ideas can we also validate these proteins and when it comes to validation then many times we are limited with the antibody approaches. So in our lab and you know there are many scientists working worldwide in the proteomics field they are trying to employ different type of platforms for doing validation. And selected reaction monitoring which is mass spectrometry based triple quadrupole mass pack based assay it is coming very powerful way of doing validation especially in the context of when we very when we do not have antibody then we can do these kind of measurement of civil peptides and try to monitor their progression in the mass pack and look at their intensity and then try to utilize that information to look for the protein quantification. Of course you know this whole slide and content is very complex but the idea is to just simply give you the feel that the same protein we are trying to measure using western blot using the protein arrays and using the you know selected reaction monitoring based targeted assays and can all of them give us the confidence that these proteins are always showing the same trend when we are trying to measure them for the healthy versus disease patients. So from this part of meningioma patient serum sample screening for auto antibody using human proteomerase I hope you are convinced that this is our very robust platform which in a single short analysis can give you idea for how many proteins are dysregulated and in this particular case we found 489 proteins were dysregulated in the grade 1 and 104 proteins in the grade 2 patients of meningiomas. Then further looking at you know the much deeper biology of them different type of pathway analysis when we start narrowing down the candidates and we found several proteins are very interesting biomarker candidates for meningiomas like IJG4, TPD52, selenium binding protein and we try to also do their validation using different type of proteomic technologies. Now this work was published in Onko Targets again the work involved variety of collaborators from Qatar Memorial Hospital the clinicians some technologies from the Johns Hopkins University and we have even within IT Bombay the PhD students post-doc rates and other faculty members who are bringing expertise from the you know the big data analysis disciplines. So the entire team is then able to try to understand the complex problems and in the last lecture I talked to you about the gliomas and this was a study based on the meningioma serum sample screening. But when we talk about auto antibody screening it is always not limited only with the serum sample. So I am now going to talk to you about of the same patients we also had for not as many as I talked for the glioma and meningioma for the serum sample but we also had the cesium sample from the less number of patients and we thought can we start using the same platform of human proteomerase and add the cerebrospinal fluid samples on the chip and see what kind of auto antibody we can see when we have the cesium samples. So just to give you the you know brief idea when we are talking about you know the screening using variety of samples one could look for variety of you know a sample type it could be the tissue biopsies it could be the blood serum of plasma or it could also be cerebrospinal fluid. So the top panel shows you the variety of options available for looking at the biomarker candidates but cerebrospinal fluid that is the most proximal fluid which is present in the CNS tumors and 25% proteins are you know very specific for the brain only. It is less invasive technique although it is not commonly used for the you know any kind of test but in the complications when they are taking the CSF sample this could be a valuable sample for looking at auto antibodies. CSF is very attractive because it is closest from the tumor location and it has a lot of advantages which is shown in the table here. It very much mimics the biochemistry of the brain it is present with the blood brain barrier low complexity and that number of proteins have been reported which are very unique to the brain only in the CSF. So in some way it is very true reflection of the physiology which is happening inside the brain but it is challenging because the protein concentration is very low and the salt concentration is very high. So to get the right signal after removing the salt is one of the major challenge of course availability of the CSF sample is not always. So one would only use that when it is required for clinical conditions and it is invasive technique not non invasive although less invasive as compared to the tumor biopsies. So I will summarize the results in a nutshell in a very brief manner. So when we looked at the GBM patients auto antibody response using cerebrospinal fluid we are now able to see some new proteins emerging which we had missed out from the serum samples and one of them interesting protein is chalurin road GEF kinase. Now this protein promotes the exchange of GDP by GTP regulates neuronal shape growth and plasticity and you know it is actually involved in the actin cytoskeletal remodeling it is also detected as the driver 1 mutation in the other type of which in SCC cancer. Other protein which looks interesting is nuclear protein 4 which is expressed predominantly in the fetal brain and testis and localizes in the nucleolus and has a RNA binding domain. So both of these proteins showed in majority of the grade 4 patients very high and robust signal from the CSF sample whereas in the control we could not see much of the signal from these proteins. And the proteins interesting is the structural maintenance of chromosomal protein 1a and this is the part of a cohesion complex and very important part of the kinetro core it interacts with BRCA1 and is involved in the DNA repair. Other interesting protein was SW1 SNF complex subunit SMARCC2 protein which is a subunit of the large ATP dependent chromatin remodeling complex it actually regulates the transcription of the genes by altering chromatin structures. So again the similar kind of workflow and data analysis strategy was employed here you can see lot more heterogeneity in the different patient samples and the signals and the pre normalization the patterns are very different but only after post normalization when all the signal looks uniform now we can see that which are the proteins differentially expressed with high confidence and the below panel the the volcano plot shows if you are not normalized you will see large number of proteins in the red which are showing the differential expression whereas on the right side now you can see the less number of protein which are showing the differential expression after doing the post normalization. The similar strategy we applied for the CSF of the meningioma patients idea is can we now find out some new proteins which were missed out using the serum sample and of course we did see some new proteins emerging again in the left side you see a heat map the comparison of the control versus you know the meningioma patients which are from different grades and some interesting protein like RBPGA protein which is the DNA binding protein and also principle effector of the Norse signaling pathway showed up regulation in the grade 1 and grade 2 patients which was totally absent in the control. Another protein which is LDL red 4 protein which is involved in the attenuation of canonical TGF beta signaling and shows increased abundance which is also associated with the increased cell proliferation and migration in the hepatic cancer cells not known for the brain tumors but shown in the liver cancer again showed you know very clear signal distinct signal in the grade patients of meningioma grade 1, 2 and 3 and no signal was seen for the control population. Then ideas can we start mapping these proteins in the interaction networks and look at where they are showing the maximal role and it gives at least some idea that you know many of the proteins which we are identifying they are involved in the cytokine cytokine receptor interactions, transmembrane receptor protein also belonging to the serine 3 unit kidney signaling pathways most of them are showing some GTPase activities also involved in the membrane regions. So you start getting some idea for the biological consequence of you know these proteins where they are showing their immunogenic properties. So from this part we can conclude that a few proteins show distinct response when we looked at the cerebrospinal fluid samples both in the gliomas and meningiomas and of course we have the panel of the protein which are quite distinct in the glioma patients as compared to the meningioma patients. Proteins like SMC1A and SMARCC2 were found to be associated with chromatin remodeling and showed immunogenic response in the higher-grade patients of the gliomas. Authentic body response for the NOL4 protein which is a cancer testless antigen protein and chalurin protein was detected in the GBM patients. Few proteins like RBPJ, LDLRAD4, EDIL3, NINJ2 they showed significant immunogenic response in both meningioma grade 1 and grade 2 patients. I hope now you are convinced that protein arrays could be a very powerful way of looking at autentic body responses from the serum and cerebrospinal fluid and one could look at different type of autoimmune diseases one could also look at even different type of cancer and complex diseases where there is a possibility of autentic body production. Now, let us be shift gear and very briefly talk to you about protein-protein interactions using NAPA arrays. So, most of the study which I talked to you was essentially on the purified proteins printed on the chip and 17,000 to 18,000 proteins were printed and those represents the maximum proteome coverage for humans, so human proteome arrays. At the same time the same kind of arrays could also be utilized for studying protein-protein interaction. But protein-protein interaction becomes more you know interesting if you think about how best you are going to look at the protein expression in its you know most native condition possible and that is what I think NAPA arrays provide a much more powerful way because we are expressing the protein from the DNA directly on the chip in situ and then think about an experiment where when you are you know the query protein for which you want to measure the interaction if that protein also is the DNA and going to get expressed along with the protein when you add the in vitro transcription translation mix then that will be very powerful. So, I am just showing you this image which talks to you about studying protein-protein interaction using NAPA arrays using co-expression analysis. Here as I mentioned idea is that on the chip you have let us say 10,000 spots which are all having the cDNA and you are going to express the protein using in vitro transcription translation mix which we talked earlier the cell-free expression based arrays. Now when you are adding your query protein rather than adding the purified protein you are simply adding the DNA of that query protein. So, in this case we are adding the DNA query for the FOS protein but keep in mind that now you have to have some strategy for the measurement of the interactions. So, now idea is if the FOS protein is you know with the DNA is going to come and bind on the chip it is going to show interaction with the June protein which is known in tractor and might be showing interaction with some other protein as well. So, how to measure these protein? So, the FOS in this case was having the flag tag whereas all other proteins on the chip is having the GST tag and based on this if all the proteins are expressing the proteins on the chip and the DNA from the FOS then now we are going to have the proteins expressed on the chip and going to show the interactions where it is going to show the binding. This is an idea for doing the co-expression based experiment. So, a similar experiment we were also teaching some students at the Code Spring Harbor Laboratory in one of the previous courses which I conducted with Dr. Josh LeBair. You have heard Josh talking about different type of applications. I am trying to convey you here that these experiments are not very difficult to perform and these experiments were done with the participants in a workshop which we conducted without having the you know kind of you know big laboratory setup over there. We took these arrays which having the CDNA printed within APA chemistry and the expression was performed in the course by the participant themselves. An idea here was the same experiment can we use you can see the layout of the chip on the left side where we are hoping that you know the FOS is going to bind to the June with the yellow spots which are seen and on the right side you see after doing the actual experiment we could see those signals for the duplicate spots of the June showing the binding with the FOS query protein. We also looked at measurement of anti flag antibody and can we now measure the where the FOS proteins are binding and based on that we could now measure the signal and look for the quantification of that. So this shows another small array which again you know idea is to just show you the crew data the small raw data which even can be obtained in a small setting in a workshop which participant themselves are able to do these experiments and in this case again with the FOS flag query DNA which is after co-expression going to bind to the June protein when could start measuring the signal and you can see these are the duplicate spots which are seen for the June protein. And likewise we had another experiment here we showed that you know now when we are measuring the binding with the FOS not only that you know the FOS protein which is you can see in the red is binding with the June but also we could identify some new interactors and in this case FKBP5 was shown as a potential new interactor where the FOS protein was binding. Additionally now these arrays could be done for use for the other applications as well especially to look for the kinase assays and Josh has talked to you about various type of phosphorylation experiments but I am just again showing you this experiment done in the course by the participants where they took these upper arrays did the phosphorylation experiment dephosphorylation and rephosphorylation. Idea was can we monitor some of these kinase proteins and look at their expression and because we do not want to have any confounding effect coming from the ribate reticuloside lysate or IVT mix that is why we are doing dephosphorylation step in between and then again dephosphorylation to ensure that those phosphorylation what we are measuring to they are actually coming from the actual activity of these kinase proteins. So after doing that then we were able to measure the signal for you know the controls and dephosphorylation, post lysate phosphorylation, rephosphorylation of these proteins and some proteins you know look quite interesting when we start measuring their trends for the ABL protein, BCR ABL mutant we had on the chip and you know just it gives you the idea that lot of interesting biology could be studied very rapidly if you know how to utilize these kind of array platform to address biological questions. So I hope I was able to convey you that you know protein arrays could be utilized for different type of interesting biological applications looking at the clinical context to that one could look at auto antibody screening looking at the you know a protein of interest one could do protein-protein interactions one could also look at you know various type of activity for the PTM modifications like kinase activity and different ways of arrays can be utilized from the human proteome arrays to the NAP arrays. So protein microarrays are definitely a very powerful platform for various applications specifically for measuring the protein protein and protein other biomolecular interactions, post-translation modifications and biomarker discovery. Auto antibody screening definitely shows the power of using protein arrays. Directly from the crude serum or cesar samples we are able to screen this kind of you know the auto antibody response which is very close to the clinical type of testing which is not possible using very robust mass spectrometry based platform. Mass spec card you know advanced instruments but for doing the analysis you need to extract the protein out from serum and cesar samples clean them up digest them get the peptides out and then only analyze the peptide then do a lot of analysis. Now when you are you want to analyze the serum or cesar you have to also do additional step of depletion you want to remove the abundant proteins and then only you will be able to get the right signal. So therefore using arrays and directly by using the crude serum or cesar sample shows you the you know how powerful this platform is you will you are able to measure the the signal which is not possible from other technology platforms. I hope you know you are learning these techniques for in the field of interactomics and big data sciences but also able to appreciate where each of the technology gives you more advantage and where the technologies have limitations. I hope you will start studying more about not specifically only protein arrays but variety of protein technologies and start thinking about which technology can address your belcher questions of interest much more powerful and robust manner. Thank you.