 |
||||
| 2008 JASN IMPACT FACTOR 7.505 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
|
||||
| 2008 JASN IMPACT FACTOR 7.505 | HOME AUTHOR INFO EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP | |||
| CURRENT ISSUE | ARCHIVES | JASN Express | ONLINE SUBMISSION | |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Special Article |
Departments of Medicine and Physiology & Cellular Biophysics, College of Physicians & Surgeons, Columbia University, New York, NY
Correspondence: Dr. Qais Al-Awqati, Departments of Medicine and Physiology & Cellular Biophysics, College of Physicians & Surgeons, Columbia University, 630 W. 168th Street, New York, NY 10032. Phone: 212-305-3512; Fax: 212-305-3475; E-mail: Qa1{at}columbia.edu
 |
Abstract |
|---|
cells secrete acid, while β cells secrete HCO3. It was found that β cells convert to type when the animal ingests an acid diet or when isolated perfused tubules are exposed to acid. This conversion of cell phenotype requires the induction of new genes, accompanied by a change in cell shape, development of microvilli, and apical endocytosis. All of these changes are reminiscent of terminal differentiation in epithelial cells. Using a β intercalated cell line, the cause of this phenotypic change was identified as a new extracellular matrix protein, which was termed hensin. When the action of hensin is blocked, the conversion of β to intercalated cells is prevented and the animals develop distal renal tubular acidosis. Hensin is expressed in most epithelia, and global knockout of hensin results in embryonic lethality at the time of development of the first columnar epithelium, the visceral endoderm. Furthermore, hensin also seems to be involved in the differentiation of transitional and perhaps stratified epithelia as well. A large number of human carcinomas have deletions in the human ortholog of hensin (DMBT1). Collectively, these studies demonstrate that hensin is a mediator of terminal differentiation in many epithelia. |
Introduction |
|---|
|
I tried to find a residency position in the United States in an academic medical center, but like so many other physicians from the third world, the only one that would accept me was a chronic disease hospital in Baltimore, which I took without having any idea what that hospital did, but luckily, it was on the same campus as Baltimore City Hospital (Bayview Medical Center today), and being one of the teaching hospitals of Johns Hopkins, I was able to attend many great academic teaching conferences. Its chief of medicine was Julius R. Krevans, who took great interest in my career, appointing me senior medical resident the second year I was there. This was the most important career push I had ever received, since it lifted me from the bottom-feeding region to one where the "real action" in medicine took place. I have fond memories of going to concerts and plays together with Juli in Baltimore and having many wonderful conversations with him.
By a stroke of luck, William B. Greenough, one of the best scientists in cholera research, returned to Johns Hopkins from Dhaka (in Bangladesh but then East Pakistan) to become the chief of infectious diseases and having heard of my work in Baghdad asked me to join him in research on the mechanisms of development of diarrhea in cholera. He also encouraged me to write up our experiences in Baghdad, and I submitted them to Lancet, which immediately accepted it and sent back a rewritten text that only now I realize, on reading the paper, that they removed any mention of glucose-coupled Na transport.4 I still remember how happy Juli Krevans was when the paper got accepted, which he mentioned at Grand Rounds. Shortly thereafter, Krevans was recruited to be the dean of the UCSF School of Medicine.
As a fellow in infectious diseases at Hopkins, I worked under the supervision of Greenough at Johns Hopkins Hospital and Michael Field at the Beth Israel Hospital in Boston, and we discovered that purified cholera toxin causes chloride secretion in the intestinal mucosa through stimulation of adenyl cyclase.5 This experience also provided me with two life-long friends, friendships that have never stopped to inspire and instruct. These studies in the short-circuited small intestine introduced me to the principles of ion transport across epithelia, and I wanted to learn more, so I applied and was accepted to work in the laboratory of Alexander Leaf at the Massachusetts General Hospital (a Homer Smith awardee), where I worked from 1970 to 1974.
Despite what I thought was hard work, I did not achieve any serious scientific result; in retrospect, nothing could rise to the level of the cholera toxin work except to follow it up. On the other hand, the almost daily conversations with Alex Leaf forced me to think like a scientist and to understand the role of the physician-scientist in the world today. He thus had the most influence on my way of thinking, but my research project was not leading anywhere, and I became more and more dispirited. When Philip Steinmetz (another Homer Smith awardee) invited me to go with him to the University of Iowa as an assistant professor, I immediately accepted. In Iowa, I met David Dawson, who, like me, was recently recruited there; David is a real "professional" ion transport physiologist who obtained his PhD in epithelial transport at Yale (with Peter Curran and Stanley Schultz). He was the first brilliant young contemporary I had come across, and he actually taught me all I know about this subject.
Steinmetz also taught me all of the intricacies of working on acid-base transport, which is the field that I think I am working in to this day, although it is difficult to discern that when you read the titles of my papers. But within a few years I became restless; Iowa City had wonderful music, and although it had the best cello teacher I ever trained under, I felt socially isolated and I started to look at many positions. One of the biggest temptations was an offer by Floyd Rector, who wanted me to take the position left by Barry Brenner on his move to Boston. I went to San Francisco, where I stayed with Juli Krevans and had a wonderful weekend with them.
But the temptations of San Francisco were countered by another offer. Daniel Kimberg, whom I met through Michael Field, had been asked to return to Columbia to be the Chief of Medicine, and it took me one millisecond to agree. I arrived there in 1977, and within a few minutes of my arrival, I knew that I would never leave. It was the environment of Columbia but also New York City that had made such an impression on me.
Within a year, I became the chief of nephrology, which lasted too long, 25 yr; I don't think that I was suited to this post not knowing anything about finances or the running of a clinical enterprise. Luckily I was able to recruit a cadre of great clinician investigators starting with Gerry Appel, who seemed happy to do his work without any interference let alone any direction. What a relief it was to have Don Landry as our new chief and to see what was missing during my tenure; our division continues to grow, and we have a vibrant set of colleagues who are happy to be there.
At Columbia in the late 1970s and early 1980s, we were very lucky in having an astonishingly large number of physician-scientists in our house staff. I have the fondest memory of a weekly meeting with a group of about eight or so house officers to discuss a paper or simply to talk about science. I learned a lot, and I know that they enjoyed the interaction, since all have remained great friends to this day. A string of incredibly talented house staff decided to work in my laboratory, and I owe all of them the greatest gratitude.
But the scientific and cultural environment of New York City is the one that had the greatest impact on my development as a scientist. New York City was and still remains the capital of world cell biology. I became a cell biologist not through rigorous training but through deep friendships with cell biologists in the city whose work was so spectacular that it inspired emulation.
So when I look back, I am astonished at the randomness of what happened to me; there just seemed to be no volition—no actual decision was made that consciously guided or changed my direction. There was some internal drive giving momentum, but it is the momentum of a ricocheting bullet in an enclosed space. When compared with the career of others, where I see so much planning and direction from inside, I occasionally get a momentary sense of regret; why didn't I go to work with this or that great scientist in the 1970s? But that feeling quickly dissipates when I realize how many great people I have met, how much influence I have received from mentors, but especially from friends whose trust and confidence taught me that a life in science is a worthwhile endeavor.
Although I never have been accused of humility, I have the example of my wife, whose wonderful work as an archeologist and restorer always put my own much more abstract works in correct perspective; it is one thing to make some modest scientific advance, but it is quite another to take ancient buildings and make them shine again, especially when this is done in regions of the world that are most in need of these monuments.
|
TERMINAL DIFFERENTIATION IN EPITHELIAL ORGANS |
|---|
|
|
URINARY ACIDIFICATION IN THE CORTICAL COLLECTING TUBULE |
|---|
form secretes H+.6 An apical Cl:HCO3 exchanger (pendrin) and a basolateral H+-ATPase mediate secretion by the β cells, while cells secrete acid by an apical H+-ATPase and another basolateral Cl:HCO3 exchanger (AE1 protein). Metabolic acidosis converts the collecting tubule from a state of HCO3 secretion to HCO3 absorption (i.e., H+ secretion).
To examine the mechanism of this dramatic change, George Schwartz did an initial "book-keeping" study where we counted the number of each cell type. Previous studies showed that intercalated cells have vigorous apical endocytosis, while β intercalated cells essentially have no apical endocytosis.6,7 We found that the total number of intercalated cells remained the same but that the number of -intercalated cells increased by a factor of 10. These studies, now more than 20 yr old, suggest that the reversal of direction of transepithelial H+ transport is due to conversion of the HCO3-secreting cell type to an acid-secreting cell. In both cell types, Brown et al.8 found that it is the same vacuolar ATPase that is located in the apical membrane of the form or in the basolateral membrane of the β cell type. The reversal of polarized distribution of the H+-ATPase attracted much attention among cell biologists, since it contradicted the deterministic idea (current in the mid-1980s) that polarized sorting of proteins in epithelia was due to the presence of targeting sequences in the proteins and hence was "immutable."
We initially had experimental evidence for the reversal of targeting of the AE1 protein (Cl:HCO3 exchanger) as well, but others were not able to confirm this finding. More recent studies showed that the apical Cl:HCO3 exchanger is pendrin (Slc26A4)9,10; hence, we have had to revise our initial hypothesis by saying that acidosis reverses the polarized distribution of only the H+-ATPase, removes pendrin from the apical membrane, and induces the expression of AE1 protein and apical endocytosis.
Despite the fact that our initial paper on the conversion of β to intercalated cells continues to be widely cited by the large number of studies that followed, reading all these papers gives one the impression that they seem to exist in a parallel universe. Even the names of these cells are different in much of the literature; Phil Steinmetz11 first described HCO3 secretion by intercalated cells and then termed the two cell types β and ,12 so why do other authors change the nomenclature to B and A cell types? When the effects of acid-base changes is studied in detail, changes in the distribution of the different sub-types are often found, however no hypothesis is proposed to explain these changes. The location and description of the various subtypes of intercalated cells are given, but the only question being answered seems to be that needed for quantification of subtype of cells. Of course, scientific progress must begin with accurate description, but one would think that after more than 20 yr something more than a catalogue of disparate cell types would have developed. Our hypothesis predicted from the beginning that the whole catalogue of intercalated cells represents a spectrum of cell types going from one extreme (β) to another ().
|
HENSIN: THE MOLECULAR SWITCH THAT CONVERTS β TO INTERCALATED CELLS
|
|---|
When these immortalized cells were seeded on filters, they formed tight transporting epithelia, but acidification of the basolateral medium did not change their phenotype. Either they lost the response, or else the response to acid was mediated by paracrine factors, as suggested by Wesson et al.15; however, an "accident" occurred in the lab when we found out one day that the monolayer secreted acid. Thanks to Janet van Adelsberg's obsessively detailed lab notebooks, we were able to trace this finding to an error in calculation, causing the cells to be seeded at superconfluent density. It turned out that when cells were seeded at subconfluent density and allowed to reach confluence, they resembled β intercalated cells, but when seeded at superconfluent density, this clonal cell line became identical to intercalated cells (Figure 3).16 But this plasticity was not simply a matter of density, since after cells were seeded at low density and became confluent, they had essentially the same cell number as the high-density seeding monolayers, yet they remained β intercalated cells. We concluded that high-density seeding acted as a developmental switch, which turned the "fate" of these cells from one cell type to another.
|
We also discovered that when these cells were seeded at high density, they deposited a protein in the extracellular matrix. Intercalated cells seeded on this protein at low density became intercalated cells (Figure 4). These results conclusively demonstrate that it was not density per se that produced the change in phenotype, rather density at seeding. Jiro Takito17 then purified this protein using a miniaturized assay for the development of apical endocytosis in low-density cells and called it hensin, finding that it is expressed in most epithelia.
|
|
HENSIN DEPOSITION IN THE ECM REQUIRES POLYMERIZATION |
|---|
intercalated cell. Purified hensin from the ECM has a very high molecular weight (in excess of 10 million), and negative-staining electron microscopy shows that it is composed of fibers each of which is composed of many fibrils. Extensive biochemical studies by Chinami Hikita18 showed that hensin is secreted as a monomer in the β intercalated cell but undergoes a complex polymerization in the extracellular space, which leads to the formation of its insoluble ECM form. Remarkably, monomeric hensin is not functional; only hensin in insoluble ECM is capable of inducing the β to intercalated plasticity. Polymerization as a prelude to localization in the ECM is a common theme of matrix proteins and is seen with collagens and fibronectin.
Two other proteins, galectin 3 and cyclophilin A, each secreted into the ECM by high-density cells, are required to form ECM hensin. Extraction of galectin 3 from functional hensin inactivates it, but the activity is regained after reconstitution of the two proteins.19 The other protein is a cis/trans prolyl isomerase likely to be cyclophilin A, the target of cyclosporin. We found that cyclosporin (or other, more specific inhibitors of the isomerase) prevents hensin deposition in the ECM and produces distal renal tubular acidosis in vivo.20 We recently discovered that the earliest step causing hensin deposition is the activation of vβ1 integrins.21 These receptors are clustered in high-density cells, and they bind hensin with high affinity and begin the process of deposition in the ECM, presumably by polymerization. Figure 5 presents our present model of the pathway by which hensin is modified and then acts on the epithelial cell.
|
|
ROLE OF HENSIN IN INTERCALATED CELL FUNCTION IN VIVO |
|---|
intercalated cells. Blocking antibodies to hensin prevented this change in polarity. Furthermore, inhibitors of the cis/trans prolyl isomerase (cyclosporin and other more specific inhibitors) also prevented the conversion of polarity, prevented hensin deposition in the ECM, and produced distal renal tubular acidosis, the conditions expected from lack of generation of intercalated cells.20
These studies demonstrate that hensin deposition in the ECM is the critical signal in the conversion of β to intercalated cells in response to acidosis. We recently obtained the most stringent criterion for this proposal when we generated mice that are deleted in hensin in the collecting tubule by breeding hensin:LoxP mice with those expressing the Cre recombinase under the control of the HoxB7 promoter. Mice lacking hensin in the collecting tubule do not have AE1-expressing cells ( intercalated cells) in the cortical collecting tubule, while there was no difference in the abundance of pendrin-expressing cells (β intercalated cells). Similarly, by real-time PCR, there was no expression of AE1 in the cortex. These studies show unequivocally that β intercalated cells convert to intercalated cells.
|
CONVERSION OF β TO CELLS IS A FORM OF TERMINAL DIFFERENTIATION
|
|---|
intercalated cells had a dramatically different morphology than β cells. Similarly, the shape of high () and low (β) density cells differed such that high-density cells were twice as tall as low-density cells in culture. The apical surface of high-density cells (like intercalated cells in vivo) has exuberant apical microvilli and vigorous apical endocytosis. Low-density cells had hardly any apical microvilli and essentially no apical endocytosis, again similar to their phenotype in vivo. When a clonal cell line was seeded at low density on a hensin-containing ECM, they developed apical endocytosis and apical microvilli, demonstrating again that the density of seeding was not actually a "contact inhibition" type of phenomenon.24,25 Apical microvilli require the presence of elaborate cytoskeleton, and indeed we found that high-density cells induce the production of villin and cytokeratin 19, proteins that are absent in low-density cells. All of these phenomena were induced by hensin purified to homogeneity. All of the characteristics of high-density intercalated cells are those of terminally differentiated columnar epithelia, and low-density intercalated cells remind one of early embryonic epithelia (e.g., primitive endoderm, kidney tubules at embryonic day [E] 12, or early small intestine), where the cells are flat and have no apical microvilli, no apical endocytosis, and no exocytosis. |
Terminal Differentiation of Other Epithelia: The Role of Hensin |
|---|
Recent results also suggest that hensin might be involved in the terminal differentiation of transitional epithelia such as the ureter and bladder. Ureteric differentiation begins when the ureteric bud starts to invade the metanephric mesenchyme. As the ureteric bud, an outgrowth of the wolffian duct, begins to divide to form the branching tree of the kidney, an early differentiation event determines that the trunk of the ureteric bud begins to look different from the branching tips.27 The trunk will eventually form the ureter, while the branching tips will form the collecting duct system. Initially, the ureteric trunk is composed of a single layer of epithelial cells. Later, the layer will be duplicated and the surface cells will start expressing more differentiated markers such as uroplakins. We have observed that hensin is deposited in the ECM of the trunk but not the branching tips. Further, when we incubated embryonic kidneys in organ culture in the presence of anti–hensin-blocking antibodies, the trunk began to dilate and branching events were reduced in number. These preliminary studies suggest that hensin might also be involved in terminal differentiation of transitional epithelia as well as simple epithelia.
|
FUTURE PROSPECTS |
|---|
Many cancer biologists believe that blockade of terminal differentiation is a critical determinant of oncogenesis, and the human ortholog of hensin (DMBT1) is deleted in a vast number of epithelial cancers28; hence, we suggest that hensin is involved in the terminal differentiation of many, perhaps most, epithelia. We believe that the study of the hensin signaling pathway will provide a deeper understanding of the process of epithelial differentiation.
|
DISCLOSURES |
|---|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
REFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J.-L. Horng, L.-Y. Lin, and P.-P. Hwang Functional regulation of H+-ATPase-rich cells in zebrafish embryos acclimated to an acidic environment Am J Physiol Cell Physiol, April 1, 2009; 296(4): C682 - C692. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W.-J. Chang, J.-L. Horng, J.-J. Yan, C.-D. Hsiao, and P.-P. Hwang The transcription factor, glial cell missing 2, is involved in differentiation and functional regulation of H+-ATPase-rich cells in zebrafish (Danio rerio) Am J Physiol Regulatory Integrative Comp Physiol, April 1, 2009; 296(4): R1192 - R1201. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
HOME
CURRENT ISSUE
ARCHIVES
JASN Express
ONLINE SUBMISSION
AUTHOR INFO
EDITORIAL BOARD SUBSCRIBE FEEDBACK ALERTS HELP |
Top
Abstract
Introduction
