The future of pancreatology: how to go where we have never been before

Abstract

The current article is part of a series of manuscripts, which together serve as a never-perfect yet candid attempt to honor the remarkable scientific stature of Professor Hans Gunther Beger. The challenge posed to the author has been to meditate on the future of pancreatology. In the following paragraphs, to the exclusion of no other, we provide an approach to future discoveries in our field. This approach combines the large scope of the new biology (eg, genomics) with the analytical power of conventional, hypothesis-driven, molecular cell biology of individual laboratories to build comprehensive functional models of normal and diseased pancreatic cell populations. In addition, the potential challenges that both the size of our field and current research funding offer to scientists focused on the pancreas are discussed. It is the author’s hope that the readers will not only meditate on the reflections offered here but, when in agreement with the positions stated, will help to implement a road map that can make our scientific field better. Never forget—the lives of our patients depend on what we do.

The writing of this article originated at the time of my visit to Germany to honor Professor Hans Gunter Beger, upon invitation of one his past students, Professor Markus Buchler, in my view the scientific successor to Hans. The meeting to honor Professor Beger was extremely stimulating at both the personal and professional levels. I have always respected and admired Professor Beger for his undying dedication to giving birth to a new generation of academic surgeons, who by sharing his commitment to excellence could push the reach of surgery to new high places. He, and the countless numbers of brilliant students he mentored throughout the years, redefined the “surgeon’s day” from a heavy “OR-loaded” profession to one that brought the bench and the microscope to a similar level of importance in their life. Perhaps his style is very dear to me because it resonates with one of my earliest learning experiences in science—that “every approach has its limitations.” Accordingly, for him, the operating room, as the only evidence-based source for his research, was not enough. Armed with an impressive intellectual power along with a special instinct for recruiting the most talented students and collaborators, Professor Beger is, no doubt, one of the founders of modern pancreatology. His team and now the team of his dearest students have been doing outstanding translational research before this concept became known as such in the Unites States. Therefore, this article is dedicated to that giant. Newton once said, “If I have seen further, it is by standing on the shoulders of giants.” Therefore, it is my hope that I can keep his shoulders close and focus on Professor Beger as a model and an inspiration for my own betterment.

I would like to alert the reader that this article commenting on the future of pancreatology will not, in spite of my own medical training, deal with surgery, but will rather aim at serving as a meditation on what we need to define, in terms of conceptual frameworks and paradigms, in order to cross the frontiers of the field that Professor Berger, his direct followers, and his contemporaries have established. Consequently, due to the limitation of the allotted space, I will focus these meditations on new foreseen frontiers in pancreatic cell biology, since as recognized by all the attendees at Professor Beger’s meeting, advances in medical and surgical treatments for pancreatic diseases cannot occur without translation of novel basic scientific concepts. For my expositions, I will also consider that, in spite of having advanced science to a new methodological revolutionary moment, we cannot proceed without an integrated theoretical scaffold that considers each of the nascent pieces of information with the biology of the cell in mind. In other words, in spite of living in an “omic” era (genomic, proteomics, etc), we must stick to the cellular theory coined by another giant, German scientist Professor Virchow. Virchow is best known for his theory omnis cellula e cellula (“every cell originates from another existing cell like it”), which he published in 1858. This theory guided his findings that not the whole organism but only certain cells or groups of cells can become sick. Consequently, sickness is the result of the acquisition of a deviant state by normal cells.

A Need to Organize Both Existing and Upcoming Knowledge Within a Conceptual Construct: The Virtual Pancreatic Cell Models

Following Virchow’s postulate of the cell theory, one can predict that the diseased pancreas results from molecular alterations that occur at the level of any of the pancreatic cells. Acute pancreatitis has at its center acinar cell injury, whereas chronic pancreatitis involves the acinar, the ductular, and the stellate cells, and pancreatic cancer, the ductular cells. Therefore, even though science is transversing the “molecular time,” the useful background of a pancreatologist should be a modernized understanding of the cell biology of pancreatic cells. In other words, we should organize the existing and newly generated knowledge to build refined virtual models for the precursor pancreatic cells as well as their adult progeny: the acinar, the ductular, and the islet cells. We must also consider major disease players that have emerged during the last decade, such as the more poorly understood mesenchymal cells, including the stellate cells. We must start organizing the knowledge derived from the biochemical and molecular datasets into cell models that inspire and guide us in our future scientific endeavors. The building of a cellular model is also essential for translating the basic concepts into medical tools, such as novel therapies. Moreover, as we build the pancreatic cellular models, we realize where our field is highly advanced and, conversely, where it is poor in investigations. For example, the best-studied cell is the acinar population, followed by the stellate cells, and lagging behind are the ductular cells. Thus, for example, the paucity of knowledge about the modern cell biology of the ductular pancreatic cells theoretically limits the data that we can translate into medical tools to diagnose and treat pancreatic diseases. In addition, it is worthy of meditation that within these cells there are molecular networks that we know by far better than others. For example, the cytoplasmic events are by far better understood than nuclear events. The cooperation, as well as antagonism, among both the cytoplasmic and nuclear network is another area under dimmer light. These areas are a few examples of new frontiers that must be explored if we want to understand how the pancreas works under normal and diseased states. Therefore, in the following paragraph, we will address what type of virtual cellular models we may build and what approaches we can use to generate the datasets needed for this endeavor.

A Two-Step Approach to the New Pancreatology: Coordinating the Efforts of the Big Group-Oriented Science and the Individual Scientist?

The term “Big Science” was coined to refer to the highly organized efforts made by the scientific community in physics to solve important projects that were of a scale far beyond what is achievable by individual scientists. An excellent example of these types of efforts is the space programs and nuclear programs in many nations. The birth of biology as a Big Science, derived from pre-existing models borrowed from physics, predicts that if we intelligently apply the large-scale approaches of proteomics, genomics, and other “omics,” we will obtain a better picture of pancreatic cancer cells than with the isolated use of tools from the past. This prediction can come true under certain circumstances, which can better be understood by briefly reviewing the efforts and resources underlying the success of the human genome project. The human genome project, the most current example of biology as Big Science, succeeded due to the remarkably large amount of intellectual, methodological, monetary, and political resources that were channeled into getting the project successfully completed. The human genome project at the National Institutes of Health (NIH) and Celera was equivalent to putting the first man on the moon by NASA. Therefore, following these examples, our community should create an international consortium that seeks to funnel similar resources to the development of functional pancreatic genomic-proteomic cell models. This is, indeed, an approach that can be the foundation of the future of pancreatology. However, it is essential to understand that (1) no scientist in isolation can put together a model of this type; (2) a consortium is better suited for performing this type of task since significant needs exist to prioritize the questions and allocate resources, which, in essence, obliges us to put most of the brightest minds together for the mission. For instance, to take a simple example, do we need more microarrays? If so, which ones? Who does it? According to which experimental and methodological design? (3) The prioritization of this type of experiment should be done with an end in mind. In this regard, an important intellectual tool is the concept of pathway reconstruction, where molecules discovered by complementary large-scale approaches (eg, combination of proteomic, genomic, and bioinformatics) can be given a distinct function within a context of a unique functional pathway. Therefore, these efforts should address questions such as how the information is processed in a manner that makes sense? What information is “noise,” and which “noise” can be information?

Hence, in my understanding, even though each pancreatologist can spend significant resources pursuing isolated experiments, which are the tools of the trade of the Big Biology (eg, microarray), a defined understanding of a particular function will not come fast enough if we do not build a Big Science strategy to solve the problem.

However, it is important to consider that the strengths of the Big Science approach are intrinsically the same as its weaknesses, namely, the rapid generation of a large body of data, which is impossible to be understood by an isolated scientist or a group lacking an appropriate guiding conceptual skeleton. Furthermore, it is essential to understand that “pathway reconstruction within the context of an appropriate virtual cellular model” cannot proceed without the fine tuning provided by the more conventional questions and tools of molecular cell biology (Fig. 1). The latter is the one that has the appropriate scale to be pursued by an isolated laboratory. The answers, however, are deeper in mechanistic insights. In other words, the route to proceed is to gain a large, yet simplified picture of molecular pathways using well-planned and appropriately directed large biology tools and refine the model by using the trade of more conventional science. The latter is also essential to add the time dimension to the biological processes such as when a particular pathway is operational? What effect and when can such an effect be expected in response to distinct stimuli? In summary, the dual approach described here can serve as a useful road map to more efficiently guide research in pancreatology in the future.

Fig. 1.

Strategy for building modern virtual pancreatic cell models. The diagram outlines the 2-step strategy to build useful virtual models for pancreatic cells. The strategy combines gathering data using a large-scale approach by means of the work of specialized consortia. These datasets can be fed to a database, which can be queried using data-mining tools. Data-mining tools are specialized computer programs with the capabilities to make sense of the data in a biologically sound context. The data could be used by individual investigators to guide their hypothesis-driven, mechanistic experiments. The data obtained by intervals laboratories should serve to refine the models; when fed back to the database, it will increase the useful information available to subsequent iterations of this process.

The Virtual Cell Model for Pancreatic Cells

The cell models we have been talking about are virtual in nature, and they can be developed following the principles of machine-learning and computer-assisted modeling (Fig. 1). There are many types of virtual modeling that we can choose as examples to better understand these concepts. The most successful example currently available is the field of molecular modeling with molecular dynamic simulation, which has been highly successfully applied to investigation of rational drug design. The basic knowledge needed to perform molecular modeling is the 3-dimensional structure of such molecules as a protein. Usually, this model is generated by x-ray crystallography or nuclear magnetic resonance (NMR) spectroscopy. The protein structure possesses a large amount of information that can be used, with the help of appropriate computer power, to predict the behavior of a similar, yet distinct, protein and its potential ligands (eg, small drugs). A protein or a fragment of it only approximately 30% identical to another protein can serve as a template. Using parameters derived from a large amount of proteins, a scientist can use the template to generate a useful 3-dimensional model of a protein of unknown structure, study the kinetic behavior of this molecule over time (molecular dynamics), and look for potential ligands. Thus, in this manner, we have derived new knowledge from a parameter-assisted virtual experiment based on the knowledge of the first protein model and the behavior of most of the proteins we know to date.

Building a complete virtual cell model is no doubt a million times more complex than the model of a molecule. However, we envision that an appropriate unit of work in the virtual cell model is the “pathway reconstruction method” (Fig. 2). In this case, a few seemingly unrelated proteins can be joined together in a pathway, their interaction studied, and new predictions of the behavior of the new pathway derived. A database that contains this information can be downloaded by individual investigators interested in adding new data and performing “in silico” simulations to reach conclusions that are worthy of being put to the test. The careful addition of appropriately curated new information can be fed back to the databank for later use, again in virtual or even real experimentations.

Fig. 2.

Example of a pathway reconstruction experiment. KLF11 targets in Panc1 cells were identified using pathway-specific arrays. Subsequently, a pathway was reconstructed using software that processes the data using semantic algorithms to search for available knowledge in literature databases (Pathway Assist; http://www.stratagene.com). The pathway depicted show interactions among proteins from a subset of proteins involved in transforming growth factor-β signaling.

Fig. 2 shows an example of a pathway reconstruction using semantic algorithms to search for known, yet not appropriately interpreted, functional connections among proteins. In this case, a series of proteins are believed to be potentially up-regulated from knowledge of the expression of their corresponding mRNAs. The modeling tool used in this situation finds the known functional relationship among these proteins and organizes the microarray result into a defined pathway. After this step, the scientist not only has knowledge about the up-regulation or down-regulation of particular genes in a distinct sample analyzed by microarray, but better, a more advanced understanding of a pathway that is either operational or disrupted by defined manipulations or disease processes. With this knowledge in hand, the scientist can now proceed to predict the behavior of the same, similar, or even antagonistic pathways and use this knowledge to perform hypothesis-driven research. Therefore, one can envision how building a virtual cell model containing the largest amount of molecules, known interactions, and knowledge of the time scale of this interaction can serve as a rapid tool to test exciting hypotheses in the shortest period of time.

Challenges in Pancreatology

The size of our field?

Having outlined a road map to successfully continue research in our field with excellence and at the highest efficiency, we have to ask whether our field is prepared to do so. This is not a trivial question if one considers that in 2007 there are few limitations in technology. Therefore, if technology is not a rate-limiting factor in our endeavors, where do our challenges lie? A significant challenge is the size of our field. In pancreatology, we can count with one hand the names of scientists dedicated to improving our understanding of pancreatic biology and pathobiology. On the other hand, as a simple comparison, we cannot say the same about neurobiology. Taking as a case study in the United States alone, the NIH has 2 large institutes dedicated to research in the area of neurological and communication disorders, the National Institute of Mental Health (NIMH) and the National Institute of Communication Disorders and Stroke (NICDS). In addition, other institutes, such as the one dedicated to understanding vision (National Eye Institute [NEI]), deafness (National Institute of Deafness and Other Communication Disorders [NIDCD]), addictions (National Institute of Drug Abuse [NIDA]), and aging (National Institute of Aging [NIA]), play significant roles in fostering studies in neurological and behavioral sciences, as well as neurodegenerative disorders. Efforts in pancreatology at the NIH represent only a portion of the portfolio of the National Institute of Diabetes, Digestive and Kidney Diseases (NIDDK) and the National Cancer Institute (NCI). Another indicator of the size of our field comes from the size of our meetings and the number of publications related to the pancreas. It is noteworthy that we pancreatic scientists meet in small groups, such as, among others, the American Pancreatic Association (APA), the International Association of Pancreatology (IAP), the European Pancreatic Club (EPC), the United States Pancreas Club, and the Pancreatic Section of the American Gastroenterological Association (AGA). Each of these meetings brings together on average 300 attendees. This number is insignificant when we compare it to the attendance at the annual meeting of the American Society for Neurosciences, which totaled more than 20,000 attendees from its 37,000 members, in 2006. These data should bring us to meditate on the questions: How many of us are doing science in pancreatology? How do we compare with other fields? What are our achievements and how do they compare with theirs? In other words, do we have the critical mass of scientists to achieve our mission, providing that we know what our mission is? The answer to the final question is that, unfortunately, we have a much reduced number of scientists in pancreatology when compared to other communities, and thus it is a great challenge to reach the critical mass necessary to further our mission.

The current funding for our field?

The more technologically advanced a field is, the more it takes to monetarily support individual research enterprises. In addition, when the playing field is leveled, and all the researchers in a particular field have access to the same methodology, time is a limiting factor in our endeavors. More time in research can be gained by choosing intelligently what experiments to do and assigning more of the best-trained individuals to perform experiments. Of course this means that “time is money.” Therefore, one can safely say that in 2007 the pace of our discoveries is directly proportional not only to the generation of innovative ideas but also to the price of technology and the appropriate economical compensation of our team members. This premise leads us to conclude that an economically rich field can make discoveries more rapidly. To illustrate this concept, I would take as an example the race for the discovery of new drugs and the underlying force fostering these types of activities. The generation of new therapeutic agents is a race for money. The best drug produces a significant amount of return. Investment in this type of research primarily comes from the private sector and is well in excess of what the government or a diseased-oriented private foundation can provide. Therefore, there is no question that the “generation of new drugs” is not exclusively but, yes, primarily a field owned by large pharmaceutical companies. With the resources in hand, new discoveries in this field are a matter of time—in particular when industry has also embraced the use of the most modern large-scale tools of the Big Biology (genomic-proteomic) and, in addition, of the Big Chemistry (rational drug design). For these reasons, not surprisingly, most of us have left the discovery of new drugs in the hands of industry.

Now let’s compare the case of drug discovery by a pharmaceutical company and the model that we use to fund discovery in pancreatology. In the United States, research for pancreatic diseases comes, as I mentioned earlier, through small pockets within the government research portfolio, as well as the generous but not extremely rich foundations and small “Pharma” for pancreatic diseases. The model for obtaining these funds is the conventional investigator- initiated research traditional in academia. This approach involves independent investigators spending, on average, 3 months writing a grant, which is reviewed 3 to 4 months after submission and, if successful, funded a year after it was written. Unfortunately, however, with the significant shrinkage in funding to academia, a grant is unlikely to be funded on the first round of submission. Some grants from our most prestigious investigators take up to 3 revisions, roughly equivalent to more than 2 years after the initial conception of the work. Consequently, the current availability of resources is extremely limited, to say the least. Pancreatic laboratories are reducing the number of individuals that we train. Trainees coming out of pancreatic laboratories are becoming more difficult to retain within our field. With fewer trainees, the time required to perform our research is longer and the acquisition of new knowledge and technology is delayed. Thus, the future of pancreatology is bleak without an appropriate increase in funding. Taking into consideration that scientists are not legislators, one should ask the question: what can we as scientists do? The answer to this question can be very complex; we can at least envision that a higher level of participation is needed. The approach should be carefully chosen since the individual scientist is powerless. Our scientific organizations must be empowered to have a voice for our representation, and we, the scientists, must be bringing those concerns to our societies. Our societies must be responsible for releasing opinion statements that can reach research administrations such as the NIH. The message must be loud and clear: an increase in funding in 2007 will not even advance but only maintain the number of investigators that are supporting discoveries in our field.

Concluding Remarks

Opportunities

At least one integrated approach can lead to a brighter future in pancreatology. This approach is to develop a large picture, a predictive cell model using the tools of the new Big Biology, and to refine it with hypothesis-driven, mechanistic, and theoretically supported molecular cell biology experiments.

Strategies

The large-scale approach requires the formation of a consortium or consortia that can define and implement the soundest investigations on problems of the highest priority in our field.

Challenges

The current challenges are the lack of a cohesive approach for forming and funding these consortia, the significant budgetary reductions in support of our research, and the current lack of representation by our professional organizations to bring our problems to the attention of the large funding entities.

In summary, the future of pancreatology is about science, not only in the daily planning of our experiments but also in the reshaping of the ways we do science in our field. The recognition that fundamental discoveries will drive the translational research and produce the knowledge and methods necessary to appropriately treat our patients and save lives is the current driving force in our field. The combinatorial Big Science–independent laboratory approach seems a sound discovery strategy for now and the near future. The challenges are to bring the most talented scientists together, to prioritize the questions that need to be asked (international scientific consortium), and to encourage our scientists to participate in shaping our professional organizations into entities that facilitate the interaction with funding agencies as well as nurture our youngest investigators.