NIH CLINICAL CENTER GRAND ROUNDS
Episode 2010-04
Time: 1:05:44
Recorded January 27, 2010
"Ciliary Diseases: Converging Observations and Diverging Models" is the topic for Grand Rounds on Jan. 27. Presenting are Dr. Meral Gunay-Aygun, staff clinician, Medical Genetics Branch, NHGRI, and Dr. Gregory G. Germino, deputy director, NIDDK. The lecture will be videocast, http://videocast.nih.gov.
ANNOUNCER: Discussing Outstanding Science of the Past, Present and Future - this is NIH Clinical Center Grand Rounds.
(Music establishes, goes under VO)
ANNOUNCER: Greetings and welcome to NIH Clinical Center Grand Rounds, recorded January 27, 2010. On today's presentation, the topic is "Ciliary Diseases: Converging Observations and Diverging Models". Presenting are Dr. Meral Gunay-Aygun, staff clinician in the Medical Genetics Branch of the National Human Genome Research Institute, and Dr. Gregory G. Germino, deputy director, National Institute of Diabetes and Digestive and Kidney Diseases.
We take you to the Lipsett Ampitheater at the NIH Clinical Center in Bethesda, Maryland for today's presentation.
GUNAY-AYGUN: Good afternoon, thank you for giving me this chance to speak in here today. I'll first start with the different types of cilia and tell you the story of the new discovery of cilia and the characteristics of human significance, including the fine cystic diseases and end my talk with some highlights with the study.
Cilia are organelles that are present on most cells of human body. And these include epithelial cells such as those lining the bronchials. Cilia has a 9.2 with nine pairs surrounding a central pair, and they are equipped with these arms that are enable the cilia to beat in a rythmic manner to propel. This is also referred to as the motile, 9 plus 2. It lacks the sensor pair of microtubules and also lacks the arm. This 9 plus 2 lacks the pair but it does have the arms. The distinction now with the classification into sensory and notes cilia, we know for instance motile cilia has some functions designed to minimize aspiration. So cilia are compartments of the cells. The cilia membrane is distinct from the rest of the cell membrane in that it is rich in receptors and monitors. And there is no ribosomes or synthesis in the cilia so the structures at the base here are important in regulating protein. And given that there are 2500 entries in the cilia today this is a highly complex process. Cilia are assembled and maintained with the evolutionary process called intraflagellar transport. There is the motor that carries it and carries the cargo back to the base. The memrane bound cargo is first loaded from golgi which are then carried up by the motor run. And then fused with the membrane and carried with the same machinery. There are 8 and many other proteins that are defective in ciliar diseases. In contrast to other organelles the cilia assembles only when cells exist mitosis into a more stationary differentiated phase. The cells that are more stationary deliver shade. One of the centrioles becomes the buddy and the cilia is built on it. And vice versa, the cilia has to be disassembled and it has to relieve the cell in order to enter cell cycle because it's become one of the centrioles of the center.
So this has implications of processes like GKD and cancer of course. Most of us are familiar with the underlying disease process of a basic defect as described by an English doctor And he realized that the sperms of these men were not normally moving. They were dismotile. And then found out that their cilia, they are lacking these motor parts of the cilia. We now refer to this group of the motile cilia as the primary dyskenesia ciliary. This is not to be confused with the primary ciliary because this is a defect of the motile cilia. There are a group of diseases that are all caused by defects of the motile missionary, if you will, of the cilia. And it's a mild end of this, the cilia might look normal under em but then the cells are cultured, they might function and still be dismotile. Like this one here is trying to move but it's dismotile as opposed to the normal cilia. Half will these patients also have the organs in the wrong place. That's because nodal cilia that has to rotate a specific angle, also requires these motors, and this is important because this is the thing that breaks the signature on the embryo because this creates this leftward flow and then down stream like a typical gene expression.
There are cycles here that deserves mention. This is the cell in the retina from here in light, this is actually a very specialized tip of the cilia. This is the connecting part, this is the base. So this is a very high maintenance cell with thousands of pigment cycles that have to go up and down every minute. So this is missionary in this connected part that has to function perfectly. So it's not surprising that then many of the protection of various types are actually components of this machinery. So the primary cillium was unnoticed for many years, and considered to be without much function. And more than a century passed between its discovery and the discovery of its function and this is significantly longer than the same time for other organelles.
The first observation that are primary cilia are clinically available came by bar. it showed that the homologue over a model was an intraflaggelar transports protein, and when this is defective, it could not build it's flagella at all. When it's defective, this is the kidney, this is the cilia. The cilia are very small. So after the cystic diseases were localized, there was evidence suggesting there is some kind of relationship between cilia the disease and function of the cilia.
The next big discovery was the recognition that the cillium is a set of signaling of the cell. So all of the important pathways including hedgehog pathways of the function of the cilia. We know these are important in embryonic patterning as well as maintaining the tissues. I'm not going into too much detail here but I mention the flow. The primary cilia are immotile, so they do not move. So in the kidney or bile duct for example, they advance and you get sense of the function of the cilia by the contribution of the protein that are require for these cells in calcium flux and down stream effects. So this is part of the mechanisms that are being discussed in the pkt. Both sides of the pathway are involved between the two and we know the pathway is important because it provides these provisional clues to groups of cells so they will know if they are cells and they will know which side is luminal and which side isn't so they'll orient their spindles accordingly. And the cell division will be in the right direction. But Dr. Germino, the second speaker actually is the person who describes these so he will tell you about that.
So the clinical features of the primary cilia, the most commonly affected organs are the kidneys and liver and sometimes the pancreas. And the mid brain and cerebellum with or without retinal degeneration are also very common. Other features include the defects of abnormal bone growth with sensory defects. The main problem with the liver in cilia diseases is not a disease of the liver cells themselves, but it is a defect of the system called the ductal plate malformation. There are excessive remnants of the embryonic bile ducts and then there is progress of fibrosis around this as opposed to normal. And continuing with the fibrosis and the disease, both are part of the spectrum of liver diseases, both are caused by the ductal plate malformation. It causes chf and if in the central part of the liver results in caroli's disease. There's a continuum and patients in general have both of them and it's called caroli's syndrome.
This slide shows the type of the movements. So the resistance and dominance types of the kidney disease are the most common forms. Patients with the syndrome, this is the old literature showing their blunted callouses and this is because they have natural pathology. Other types of kidney includes multicystic dysplasic kidneys. I have an ivp of them with a sponge kidney there because this is poorly defined and there are no genes identified at this time but i think some of these patients are likely to be ciliary diseases.
So this PKD, is the most common one. Both child and adult versions are related to kidney disease. They both have limited movements but the limited movements in ARPKD are more serious in that it results in complications where there is none in ADPKD. They both cause problems in the liver but they are from the ciliary system. Nephronopthisisis is a descriptive kidney diagnosis. It can be isolated as part of the syndrome. The main thing that differentiates this from PKD is the numb nuns of the fibrosis and the cysts are secondary. This still does not account for a majority of the patients.
Related disorders are the heterogeneous group and their combining features include abnormality of the connection of the cerebellum to the mid brain. These children, their main problem is apraxia and speech apraxia and cognitive function is spared because they have a wiring problem. They have liver, kidney involvement and eye involvement. This is also very heterogeneous. The main features of this cilia disease are cognitive defects, obesity, hypogonadism, and polydactylly. These mostly localize to the basil body or are around the central zone. These patients also have side defects some of them. Some of these proteins are also important for nodal cilia. At the most severe end of the spectrum is meckel syndrome. We know now that all of these patients have hepatic fibrosis and some this degeneration. This is very heterogeneous. The thing about these genes is they are also much more important fundamentally to build the cilia. These cannot make cilia, it's all consistent.
This slide, to summarize the liver and kidney, i picked a particular moment to keep it separate because it is outside of the chf spectrum. These originate from these so-called bile ducts, and all the others, you know, diseases can be officiated with some degree of congenital fibrosis. And the milder end of this is patients with urinary concentration defects and normal appearing kidneys. So these are somewhere here at the part of the spectrum.
The pneumo cystic disease are seen in some patients. This is what happens to the cell where the protein locks down to the cytic skeleton of the cell. So these syndromes are really rare but because they are caused by this long lists of proteins with fundamental functions in the cell, they are really likely to teach us a lot. And also give some insight on the cause on of more common problems like obesity.
So we had this ongoing natural history study on neuropathies since 2003 and we have several specialists studying in that group. We are seeing more of the syndrome in patients that also has a large number of unknowns.
So on the ARPKD mode, we're able to define the full spectrum of the kidney. So this is the most typical known part of the spectrum. And this mild one with normal size kidneys is the part that we uncovered and that's a special number of patients actually. And this is diagnosable only by hybridization ultrasound. We also found that there is a correlation actually between function and involvement of imaging.
Then we divide the patient into early presenters and late presenters. And those who become symptomatic around the time of birth require kidney transplants earlier than the others.
We look for correlations between the size of the kidneys, the volume and the preliminary function in the pediatric group, maybe we found some correlation but this wasn't as good as the dominant PKD.
We also found on completing studies that a large number of these patients had the x-ray of the ducts and large gall bladder.
We also found out that this liver is abnormally shaped liver from birth with a large left lobe extending all the way back to the spleen here. Then you ask the question whether there is any relationship between the severity of kidney and liver disease. We found out that these are independent of each other.
We also had some patients with some overlap features of ARPT syndrome and we identified in patients with one of the genes. This gene, the nations have the cilia part is caused meckel syndrome, and on the transmembrane part causes joubert syndrome. And we identified this in the loop.
We had joubert patients with kidneys indistinguishable from ARPKD. This is another interesting joubert patient where she has confirmed joubert with two good mutations in one of the genes and has a multicystic displastic kidney.
We look at the number 5 gene but we didn't find anything there. So this extended the spectrum to asymmetrical but serially.
We also had patients who were diagnosed initially as having APKD later on. We saw only two type one patients, two adult females who had the typical digital findings. The strikings things we found were the abnormalities that were similar in these patients. This is almost like duplicated. One of them had a pancreatic cyst. So these pancreas findings are less well recognized parts of this syndrome. We have several patients combinations of PKD and chf and the so-called dilated spaces that we confined in the patients next time. I think the developmental brain represents a continuum and at the mild identify of this continuum there will be some of the patients with these so called variums. These were described based on their most common findings and they have single gene diseases. But as we understand, they are genetic. The difference between them are blurbed and we recognize more and more overlap. So is this because it was explainable by way of the mutation in the protein or is it because of the severity of the protein or is this because there are modifiers in that protein complex, the functional subunits of the cilia, if you will? So i think as we understand the multisubunits protein complexes, the function of the subunits in the complex machinery of the cillium, we will sort these out better. We might collect these in the future as ciliary complex and such. So there is a broad range of clinical futures regarding defects of the primary cilia. They don't necessarily have a team. It's not like defects of mitochondria. These are not like this. This is more than an organelle, a very complex compartment of the cell. I think we have to understand what are the complexes as the protein working together and some of that we are beginning to understand.
And there are of course cell types with differences among cilia with this variation. And these cilia genes have, we know a lot of them have non-ciliated functions. And in the PKD, we'll discuss some of that. So a dysfunction of the primary cilia may affect tissues to different degrees depending on which protein is defective. And also the degree of the dependency of each tissue to that subunit of the cillium.
So I want to thank to all of the contributors here. I listed the NIH group on the site. Those who made this all possible. It was very important for protocol. And all of the other nih group, patients on families and doctors.
[APPLAUSE]
>> Okay, great.
GERMINO: Thank you. I'm going to start off with just one brief disclosure. My wife is a site director at Johns Hopkins for a multiclinical center for patients with PKD that is sponsored by pharmaceuticals and i was a so investigator until I joined NIH this past year. Learning objectives are as stated here are learning about PKD and find the latest findings and pathogen sus and try to explain the implications of the latest findings.
As any student of medicine has learned, the kidney is probably one of the most elegant examples of form and function being intimately tied together. And structure and function tied together. For the tubules to process 140 liters of fluid a day and then to process that into one liter of urine, it must have the activity of cells from the glomerulus down to the collecting duct. The cells are heterogeneous in structure and function. It is essential that the tubule have the right morphology. Too wide a tubule will result in inefficient processing of the filtrate, of course it's too narrow, you have a back flow of appreciate and you have a decrease in filtration.
One of my interests as a nephrologist and immunobiologist is to understand the factors that regulate the structure on the kidney because this is essential to understanding the kidney's function. One of the best ways to study that, i think, is to be able to actually approach it from using naturally acquired examples of processes that are resulted from defective tube genesis. Cystic diseases in a class this is fit very nicely.
First off they have a genetic origin so we can identify the primary mover and principal event and then move backwards and forwards to looking to actually the consequences both in terms of structure and in terms of functional consequences for clinical disease. They provide unique scientific opportunities. Moreover, particularly for me as a physician, it's important that they be clinically important.
ADPKD is in fact one of the most common single gene disorders affecting one in 500 to one in a thousand. It accounts for four to five percent end stage kidney disease in the United States. Unfortunately there are no good therapies at present. My lab has been studying this for the last 20 years, both ADPKD and ARPKD, the recessive form of the disease, you just heard about by meral. I'm going to talk previously about ADPKD in the brief time we have.
So one of the questions of course is can we prevent this problem. This is of course the most dramatic example that I think I’ve ever seen of ADPKD kidney. This is a 43 pound kidney taken out last year at Johns Hopkins. One of my wife's patients. This is an extreme example but most people with ADPKD have a kidney with half a liter in size. This is a significant source of morbidity and premature mortality and go on to cause end stage kidney disease. Can we prevent this? Can we also prevent these other problems.
As we heard from meral, gi manifestations are extremely common including ADPKD this is a scan of a woman we've seen at Hopkins years ago and we had the cystic liver. The cystic liver had already been decompressed. She already had removed half her liver.
Pancreatic cysts are also common. Vascular abnormalities are probably the most feared manifestations with aneurysms occurring in about seven percent, two to three fold higher than the population as a whole and they cluster in families. They happen in young age to be the first presenting manifestation. Somebody who drops dead with a ruptured subarachnoid aneurism. To understand these problems and more importantly to be able to develop therapies for these problems which are very slowly progressing, typically starting in utero and go on over the course of 40, 50 or 60 years before they make these manifestations, it's clear we have to have a good understanding of the underlying biology because this is not a rapidly progressing process in general. And so to understand this, we fortunately can take advantage of the genetics of this disease as a starting point. And so with ADPKD it turns out it isn't a single gene disorder it's a genetically heterogenouss disorder and involves in one of two genes. PKD one which accounts for 85% of the forms and codes of 14,000 base pair mrna and large protein in 4,000 amino acids. PK2 accounts for everybody else. Smaller nrna and smaller protein. The reason PKD1 is more common, the gene is larger and gc rich and more mutable and has elements that enhances its mutability.
One of the first things we could do once we had the gene in hand was try to understand the mechanism of disease. It's an autosomal dominant disease and therefore there are a number of different ways that heterozygous mutations could induce a disease process. It could be a dominant acting gene effect, it could be a dominant negative effect, it could be a haploid insufficient effect or a molecular level.
Actually studying in the 1970's, we micro dissected early APKD kidneys. We found that the cysts were focal. It was something that happened early. In a series of studies i won't go through today, we and others subsequently defined molecular biology underlying this and that is disease is actually two hit in its basis. The individual tubule tells have a mutation and have provided susceptibility and provided the previous normal copy and allows the cyst to grow. This is verified by clonal analysis and direct analysis of individual cysts. The prevailing model is ADPKD results from too little of PKD1 or 2 activity in individual cells. In fact, probably the single most important valuable in determining the progression of disease is a somatic mutation rate. The timing, the position of the second hit are important and you might imagine any factor that could influence the rate of somatic mutation could also be an important factor in terms of determining disease severity. These could include genetic factors and specific haplotin. This also explains why 1 is more common but it's also more severe in its presentation.
If you look at individual patients you can't distinguish a PKD1 affected individual from a PK2 infected individuals because they are completely overlapping. If however you look at a population, they segregate nicely with pk1 clustering with its presentation typically before age 30 with end stage kidney disease by age 50 and PK2 presenting with cysts in 30's and 40's. And end stage in their 60's and 70's. This is formed later because of a lower somatic mutation rate.
So what do we know about two proteins encoded by these genes and the pathways which they signal. Regrettably, we know much less than we need to and much less than we like. I'm not going to spend time going into this today but simply to show you one quick cartoon that summarizes what we know. Poly cystic one is a very large protein. It has 3,000 amino acids, a domain and a short c terminus when it interacts with poly cystic two. PK2 gene product is a family member of the trip p family of proteins. It's a nonselective cation channel that has preference for calcium. We have these from a receptor channel complex process. Here is a subset of all the different gene products known to bind to PKDS 1. We know them to be true with various degrees of certainty. Likewise here for PKD2. The four and one of the most recent ones to be associated with PKD2 and these are thought to form a channel complex which actually respond to flow by having calcium release. I'll come back to this in a few minutes near the end of the talk. The protein complex as we heard from meral is actually, described as being in multiple places within the cell. The first location was put at the basil membrane and in fact many of the domain structures on PKD1 look like they would be adhesive and involved in cell matrix of potentially cell interactions where it was subsequently placed. The most recent localization has been placing it in the primary cilium where it's thought to be acting as a mechanosensor for flow. A whole host of cystic related proteins are related to the primary cilium and that in turn has been thought to act as a flow sensor. This flow then triggers calcium influx which modifies tubular structure and function.
The excitement of the subbology community to the cystic relationship with cilium can't be overstated. The cell biology meeting is really a major theme in the community trying to understand the relationship between the primary and all of these various signaling proteins and all of these various cystic proteins. The model as I said in PKD is that luminal flow regulates the tubular morphology. It's been reported that pc1 functions as a mechano sensor. It activates poly cystine 2 which causes calcium influx with the 2 channel and then the ciliary signaling through unknown mechanisms suppresses growth, induces difference race, maintains tubular morphology. In some models not specifically shown in APKD but other models it's been reported that the signaling serves to turn off the cannonical pathway which is very important. In the process of turning off the conical pathway it allows the non-conical pathway to be switched on. And this then is thought to regulate pcp or planary cell polarity. As you know, when we think about polarity we think of basal lateral top and bottom. That's common thing for epithelial cells. In fact as you think about a three dimensional structure cells not only have a top and bottom, they have their neighbors and have to orient themselves with respect to the three dimensions with respect to the plane of cells. So we can set this up properly. With respect to the tubules. They have to lengthen in all directions as illustrated in the next slide.
We had postulated a number of years ago as kidney development of development as i said as the tubes are lengthening, somehow these cells have to know the front and the back. And the right from the left. And presumably some pcp-like mechanism is regulating this process. And this process can occur through two different mechanisms. One would be through oriented cell division, that is as these cells are under going mitosis during this developmental phase, the mitotic spindlessor are oriented parallel to the lumen orientation to the long axis of the tubule. As the tubes divide the tubule stretches out. The second model is the cells with a divide in any random axis but would intercolade. Under cystic diseases, you could have non-direction cell division. So instead of having the axis lined up as long axis of the to be annual they would perhaps be randomized. Might see a random expansion of a tubule size and expansion of tubule length. The second possibility is it has nothing to do with oriented cell division. It could be that intercalculation does not occur properly. So instead of moving and migrating they in fact again stay in their disoriented position and the cells, and the tubule structure just expands out. I'll come back to this at the very end of the talk. It's what may or may not be happening in PKD. There is little doubt that one and two are essential for tubular integrity. The question is how and what are they doing.
I want to describe simple experiments we've done that pose more questions and answers, but actually have important clinical developments. So we've used the genetic approach since it's hard to argue with genetic methods, generally as opposed to using antibodies that have promiscuous binding sites. We used a conditional model of PKD1. We used locked p cites flanking exons 2, 4, 5. Using various times, various different recombinates patterns you can turn it on and off. Using a balance creator which causes random activation, you can produce a model that looks very much like human ADPKD. Using other inducible models we can ask what are the consequences of delayed inactivation. Can we separate out the development role from the adult role of this protein. And so there is this series of studies I’m going to briefly describe.
We wanted to ask what happened if you take it out during the early stages as the kidney's still forming. In mice that goes up to about day eight post birth. Unlike in humans with nephro genesis is finished by third trimester of life. Mice are different. They're born with still developing kidneys. Second phase, looking at what happens during adolescence. Still reasonably rapid growth but not a developmental phase as we've seen here in embryonic phase. Last phase, is there any protein in the adult animal. If you knock it out on the P2 and you look at it 17 days layer, it's normal. You can see of course grossly severely cystic kidneys. If you look close up you can see virtually all nephron segments are equally affected and if you use markers for specific kidneys, you can see that again all nephron segments are involved. Going further to the P21 this is a three week old mouse an adolescent of sorts. Going from childhood to adolescence. Here at this point to our surprise, we looked at three weeks after birth. In fact, three months post induction, the animal's kidneys look completely normal. You look however at five months most induction it becomes cystic and by six to seven months the animals are all dying of their renal complications. We looked at what happened to adult animals. Again looking at three months, this is the cystic, this is the induced animal. You can see the kidney looks normal to show that we in fact had induced successfully and got inactivation of the gene.
We used dna testing. You can show you cut up the kidney. You can see the presence of the induced band showing in fact did achieve successful recombination and deletion. If you wait now, again six months, you see rapidly explosive widely spread cystic disease. Again, affecting all nephron segments.
As previously described in the literature, and then lost sight of for a few years as we'll talk about in a minute, cystic disease in mice and in humans derives from all nephron segments whether it can develop early in life or develop later in life. This is important with clinical implications as well as I’ll mention in a second. Having defined this unit of time that there is at P21 the kidneys are not cystic and before that they become rapidly cystic. We did time in act valuations going backwards and forward, from P2 and backwards from P16. What I want to point your attention to was this time point right here. We found that anything before P12 that was induced rapidly develops cystic disease with a week. Anything at P14 and later that was induced, did not develop cystic disease for many many, many months. It's a two day interval. In fact there's a one day interval that determines tremendously the sensitivity of the kidney to an activation or loss of PKD1. So we wanted to know what was happening in this two day interval. And so we looked at some of the samples here. We had to kind of characterize this process. The first observation was in contrast to the clinical literature which suggests that hyper pro-liferation is – looking at this animal model which is in fact highly controlled, in fact proliferation rates are if anything only mildly higher. They're not dramatically different. We use two different markers of proliferation to look at this process. So what is going on in this two day interval. I showed you the proliferates are not dramatically higher at a time the cysts are forming. Nonetheless there are some relationships to proliferation. That is if you look at this two day interval the proliferation rate is abruptly dropping. There’s something changing in the kidney at this time. This has some relationship to the onset of the cystic disease, response to the cystic disease in this early onset model.
Delving a little bit further trying to understand this process we did of course a micro ray analysis on the normal kidneys in this process that bracket this window of time. What you can find here on multidimensional scaling is the P11 and P12 kidneys cluster together the p14 and 15 kidney gene profiles cluster together. You can see the series of genes that switch on and switch off during this two day interval. Identified as some two day time frame. It's a late stage maturation phase that determines kidney's affectability to loss of PKD1.
And so to summarize briefly, again if you lose the gene in its first couple weeks of life before p14 in mice, very, very rapid onset of disease. After that it takes many. Many months suggesting that maybe different biologies for this protein, different functions for this protein in these different developmental intervals. And this may correlate with what we see in humans as well. We see individuals with very, very rapid and very early onset cystic disease. We see on the average though people develop it very gradually over the course of many decades. My suspicion is that with the early onset disease in fact are ones that had fallen into this class. And had very early rapid second hits and then those became very rapidly cystic and grew rapidly. Whereas those later with the later onset disease had a much more gradual presentation. In humans this is again almost immeasurable. So what did we learn from this model? Rate and timing determines rate of severity. Lifetime stage determines the rate of pk1 loss. Nephron segments can be cystic if PKD1 is activation through no sparing of any segment. What is PKD1 doing in these different stages of life? Also the question that we're really intensively looking at right now are using a variety of different kind of methods to really drill down and see if we can identify the first signature, the very first sort of effect of the big bang when the gene is turned off before we see the significant and sort of severe manifestations in late disease. Certainly differences in the rate of proliferation between an adult organ and early organ are playing some role in the progression of disease in these models.
The adult kidney proliferation rates are extremely low as opposed to as I showed you in the early stage is where it is very high. The relationship between proliferation again and -- inactivation is not straightforward. Two days after the four days after the interval, it took six months to get cystic. Yet proliferation rates at this point are still many fold higher than they are in the adult kidney. So it's more than just proliferation making the kidney susceptible.
Moreover, if you look at the late onset model, the cysts form spontaneously throughout the organ all at one time as far as we can tell. If it was related to proliferation every time the cell cycles in the kidney which is happening in a random rate within the kidney, that would trigger the process and you get a cyst forming. And that is not what we see. You see the relationship again is unclear.
The other point I want to make is not to say that pro-liferation is not important in causing ADTKD you can't have some kid nation without proliferation. It is obviously a factor, but I argue it doesn't necessarily have to be a massively proliferative process to get the end stage we see. I show as an example my kids - if you look at them one day apart they don't look different. If you look at them over a force of five years there's a tremendous difference in size and other properties. Of course any parent knows this. So small changes over time can make very big differences in terms of final outcome.
And so proliferation is a factor but it's not I don't think necessarily a major targetable one. This poly cystic one as the flow kidney which is the prevailing model, how does this property relate to cyst formation. Well it may be a flow sensor in the kidney but I think if it's a flow sensor in the kidney it probably can't be doing it in the adult organ because you would expect if it was flow sensitive in the adult organ, you much more rapid response to loss of PKD1. In fact, someone who has done a similar study using one of the ifp proteins that control cilium formation do the same thing. If you knock out an ift protein, it takes, in an adult, months or a year before you begin to see cysts. In the adult organize cilium probably isn't a flow sensor.
The second observation is that I told you before trip V4 has been shown to form a channel complex with poly cystic two. This acts as a flow sensor. Yet if you knock out trip v4 in mice and this has been done, these animals do not become cystic at any point in their life span again suggesting that flow sensoryper say even in the young kidney may not necessarily be the major role of the cilium or PKD proteins. I can't exclude the function that may be important to the final phase in using a different flow symptom pathway during development than trip D4, nor can i rule out the fact that this may be acting as a kind of tonic signal so it's loss of this kind of tonic suppressive signal over time that's integrated that somehow causes a change in regulation of tubular morphology.
Just a few last observations.
Proliferation is a factor but the rate is not likely as high as believed. The flow of poly cystic one of the kidney is questionable. I can't rule out a role for poly cystic one in helping cells to reestablish their orientation after cells divide. So there's a model in zebra fish where pcp proteins have been shown to be important in regulating how neuro blasts development occurs. If you slow down proliferation in these neuroblasts having a pcp protein, bangal 2 or bangal one, these cells are able to still form, and develop normally. By slowing down it allows the cells to reestablish using other sensing systems than orientation. And perhaps PKD proteins may play a role in helping cells find the orientation after cells divide. But again, how this relates to this formation I think is still somewhat unclear.
I mentioned earlier that these PKD proteins sensing, whatever it's doing with its flow or otherwise is thought to perhaps act as a signal from going from conical or non conical or pcp pathway. One prediction of that is you'd expect to see perhaps a cell division in the normal tubules. In fact one group has looked at this using clonal analysis and using mitotic spindels he has reported that in fact oriented cell divisions does occur normally in renal development. He also suggests that in two different cystic models that this process is disregulated. He looked at a hitching of one mode annual model and he looked at a rat model which is a mutation of the human gene. In both models he shows in this paper that the axis of orientation is in fact not proper, not correct and so it would suggest that maybe ocd or oriented cell division is in fact an important cause a contributing factor to cystic disease.
In another paper by a group in London, by a group in England had a variety of other models. They looked at them and it's important to note in the heterozegous state these are pheno typically normal kidneys and they reported that oriented cell division was regulated in all of these models. This observation is curious. If oriented cell division is supposedly driving cyst formation, how would you then explain a normal kidney. In this particular example of ARPKD rat cystic again is a phenomena weeks after it's completed. If oriented cell division is an important factor it doesn't really fit the model as proposed. And in fact, just recently just on-line this last week was a paper from a group from Yale looking at oriented cell division again at a variety of different models and he found completely different results. He found again in a PKD1 mouse model he found and reports this disoriented cell division. This model does not get cystic disease in the kidney. So it's again disconnected from cystic disease. You look at PKD1 and 2. These are models where the genes have been turned off but they're not yet cystic. And again in these tubules and he shows the proteins missing. In these models, there's still no oriented, no disoriented cell division. So the link between oriented cell division and cystic disease I think is extremely unclear at this point. I might also suggest the methodology is complicated and I think there's also problems.
Is cellular dysfunction the cause of the PKD? I think it's clear that most have been localized. One important point to keep in mind is that most proteins have been localized to subcellular locations. PC1, 2, many of these are in other places within the cell.
Over 10% of the genome have proteins have been localized in one methodology or another to the cilium. So saying something is localized to the cilium is not extremely distinguishing at this point.
I think we have to ask the question how do we distinguish between the functions of each subdomain. If it's at the primary cilium and the matrix how do we know it's actually the most important partner cause of the disease process that we see.
And finally, there was a paper that came out last summer that showed that in fact if proteins though to be exclusively related to cilium formation, in fact has been shown to function in cells that lack cilium suggesting they may have other functions as well.
So the relationship between cilium and function I think is still somewhat in the air.
So the cilium model disease provides a unified means of explaining a widely divergent group of diseases as we heard. The mechanistic link between cilium and disease state really remain undefined. It's unlikely to be a single process explaining sist formation.
Finally, for clinical implications, I think the pathogenic disease and early onset versus late onset models that we use for preclinical testing may differ. All of the preclinical studies done to date for PKD have used these onset models and we have to wonder how predictive these would be for human ADPKD which is in general a late onset disease process.
The other important thing is nephron specific therapies are targeting the D2 receptor at the collecting duct, which may be less effective than hoped. We need to use the proper tools to actually have these clinical tests. This is a problem we often have in medicine using the wrong test to define our therapeutic approaches and we go back to humans and we discover to our surprise that in fact therapies are not effective. It highlights that one more time.
I want to also acknowledge the people at Hopkins with whom I have worked, those who made our mouse models, and who have been doing our detailed analysis of this two day interval.
I thank you for your attention.
[APPLAUSE]
Are there any questions?
QUESTION: The question is why do individuals with ARPKD have different severity of kidney and liver disease. They can have very severe liver disease and mild disease and vice versa. Is that seen in ADPKD and do we have any mechanistic explanations.
GERMINO: Two points. One is that i think an ADPKD we made a determinant of the severity is that somatic mutation rate. So i think that really is a variable that is affected by a lot of different things but i think that probably is a wild card that makes it why you could explain why you could see differences in terms over organ manifestations. The second point as to ARPKD is that i think there are some rough geno type/ pheno type correlations. When you talk about a single individual and why they may have different liver and kidney manifestations we don't really understand that. If you look at different mouse models there are different mouse models, four out of the six have no kidney manifestations. They just liver manifestations. Two out of the six have different manifestations. We have a model just like humans but it was later onset. And by changing the strain. So it is preserved.
Thank you.
[APPLAUSE]
ANNOUNCER: You've been listening to NIH Clinical Center Grand Rounds recorded January 27, 2010. On today's presentation, the topic was "Ciliary Diseases: Converging Observations and Diverging Models". Presenting were Dr. Meral Gunay-Aygun, staff clinician in the Medical Genetics Branch of the National Human Genome Research Institute, and Dr. Gregory G. Germino, deputy director, National Institute of Diabetes and Digestive and Kidney Diseases. You can see a closed-captioned videocast of this lecture by logging onto http://videocast.nih.gov -- click the "Past Events" link -- or by clicking the "View Videocast" link on the podcast homepage at www.cc.nih.gov/podcast. The NIH CLINICAL CENTER GRAND ROUNDS podcast is a presentation of the NIH Clinical Center, Office of Communications, Patient Recruitment and Public Liaison. For more information about clinical research going on every day at the NIH Clinical Center, log on to http://clinicalcenter.nih.gov. From America's Clinical Research Hospital, this has been NIH CLINICAL CENTER GRAND ROUNDS. In Bethesda, Maryland, I'm Bill Schmalfeldt at the National Institutes of Health, an agency of the United States Department of Health and Human Services.
