The convergence of science and technology in Mississippi: I. advancing the frontiers of biomedicine (1).

Science and technology have always been closely interrelated, and major advances in one area are intricately related to innovations in the other. Over the past two decades, scientific and technological breakthroughs have dramatically revolutionized the broad field of biomedicine, and continued progress is anticipated in this field for the foreseeable future. However, the generation of these exciting technical advancements demands a significant and sustained investment of intellectual and economic resources. Current data suggest that a "critical mass effect" leads to significant regional variation in the creation of scientific discoveries and technological innovations. These data also illustrate that scientific endeavors in Mississippi have not yet attained the level of scientific productivity associated with leading scientific regions of the nation. We suggest that improvements in the quality of scientific activities in Mississippi will lead to economic improvements and to improvements in Mississippi's quality of life. In this article, we will discuss two major research initiatives in Mississippi that are creating opportunities for unprecedented scientific and technological advancement in biomedicine. The National Center for Natural Products Research (NCNPR) serves as a model for building a research center that can capitalize on pre-existent scientific and technological expertise. The University of Mississippi Medical Center Cancer Institute (UMMCCI) presents an opportunity to build a nationally competitive research program in an area in which Mississippi must strengthen its scientific expertise. We propose that intellectual capital is the most important investment for the success of both the NCNPR and the UMMCCI.

IS THERE A DIFFERENCE BETWEEN SCIENCE AND TECHNOLOGY?

Anyone who has ever judged a school science fair knows that the line between science and technology is often blurred to the point of invisibility in the minds of the general public, yet a clear distinction exists between the two. Science is a philosophical approach to understanding the underlying basis of reality. Technology is the means by which humanity successfully manipulates reality.

History has shown that progress in either science or technology can lead to unanticipated outcomes. Although some outcomes have occasionally been detrimental, it is generally accepted that advancements of science and technology lead to an overall improvement in the quality of life. Scientific and technological advancements are indisputably responsible for the dramatic extension in human life expectancy, which has linearly increased by approximately four decades over the past 16 decades (Oeppen and Vaupel, 2002). Science and technology have generally advanced in tandem, and these advances often profoundly affect society. For example, Galileo's technological breakthrough, the telescope, allowed him to glimpse four moons of Jupiter (Galilei, 1610). This observation inspired his outrageous and enduring scientific proposal of heliocentric orbits, overturning thousands of years of belief in geocentric orbits in our solar system (Galilei, 1632).

How can one plot the course of science and technology? The metrics that gauge the success of technology account for the fact that technology is a tool. The assessment of success versus failure is simply based on whether or not a given technology can accomplish the task for which it was designed. If this primary assessment shows that a technology is successful, then secondary measures of relative success are assessed through other questions: How efficient is the technology? How effective is the technology? How convenient is the technology to use? How accessible is the technology? How reliable is the technology?

In a free market economy, these secondary questions are reflected in the market demand for that technology. There are rare exceptions in which market demand may not accurately measure technological success, such as technologies which are governmentally restricted, technologies with highly specialized applications, or technologies which incur prohibitive monetary, societal, or environmental costs(8). Historically, technological success has been closely tied to wealth creation, at both individual and societal levels. A successful technologist, such as Thomas Alva Edison or John Craig Venter, often becomes a wealthy individual.

In contrast, the measure of scientific success lies within the essence of empirical scientific philosophy, namely, the testable hypothesis. The scientific success of a hypothesis is based on its capacity to predict an experimental outcome. In scientific disciplines unsuitable for experimental manipulation, the success of a hypothesis is judged by its capacity to consistently explain a broad range of observations. Since the successful scientist is rewarded with knowledge, but not necessarily with wealth, one might have to rely on non-economic parameters to measure the relative scientific success of individual scientists, scientific organizations or geographic regions.

So can we "measure" biomedical science and technology, and can we "measure" the relative success of Mississippi in these areas? Can we identify strategies for growth, improvement and development? What benefits might one expect if Mississippi became an international leader in advancing the frontiers of biomedicine, as judged by objective metrics?

It is difficult to directly count the number of innovative scientific discoveries or the number of validated hypotheses generated by all scientists, and nearly impossible to find any single method that yields results that can be compared across disciplines. Various agencies have compiled a large set of indicators to provide an assessment of the "scope, quality and vitality" of the scientific activity within a given region. Some of these indicators are shown on a state-by-state basis in Table 1. One data source is the biennial issuance of the "Science Indicators" report (National Science Board, 2002), a comprehensive document providing statistical information on elementary, secondary and higher education, technically skilled workforce status, private and public funding for scientific activities, and other indicators of U.S. competitiveness in the broad area of science. The National Science Foundation's Division of Science Resource Statistics publishes other reports summarizing state-by-state data from various sources (Payson, 1999; National Science Foundation, 2003). The Mississippi Innovation Index (Mississippi Technology Alliance, 2004) provides an objective measurement of Mississippi's innovation-related activities in comparison to its neighbors in the southeastern USA. The Mississippi Innovation Index is computed from a matrix of parameters such as wealth creation, technology business development, statewide research capability, industrial productivity, university research and development activity, technological workforce development, business research and development activity, as well as investment capital.

The data examined in the National Science Foundation studies and the Mississippi Innovation Index relate to both research and development, which intermingles both scientific and technological activities. Because these studies primarily focus on economic data, they are more reflective of technology rather than science. In order to measure the quantity and quality of biomedical scientific activity, we also sought data from the scientific peer review process itself. This process is designed to assess objectively either scientific merit, in instances of research publications or of funding proposals, or scientific accomphlishment, in instances of awards or honors.

One way to measure regional research productivity is simply to count the number of retrievable scientific publications in the PubMed database (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed). Just as a prospective employer is cautious about evaluating only the number of publications, but not the substance of these publications, when evaluating an individual scientist's curriculum vitae, so too should we be cautious when utilizing PubMed citations as a metric tool for assessing regional scientific performance. Statewide publication tallies derived from the PubMed database are presented in Table 1.

In an attempt to quantify scientific productivity, one should also attempt to quantify the quality of the scientific product. Most scientific databases are not designed for the purpose of regional comparisons, so any individual dataset should be viewed cautiously. The PubMed database was re-examined, and only the manuscripts from a handful of "high profile" journals were counted (Table 1). These journals were Science, Nature, Cell, The Proceedings of the National Academy of Sciences, and The New England Journal of Medicine. It should be noted that some would view these as the highest profile science journals out of a field including several hundred journal titles and that some of these five journals do not restrict their content to the area of medical science. To visualize the statewide distribution of "high profile" scientific output (as a surrogate for high quality scientific output), we replotted these data in cartographic fashion. Because of space limitations, only data representations that are statistically meaningful (see below) will be shown. Figure 1 is a map showing the statewide distribution of "high profile" publications per total PubMed citations. This is a rational basis for normalization in order to reflect the quality of scientific output, and as will be shown below, these are two strongly correlated parameters.

Occasional problems arise from regional analysis of the PubMed database due to inaccuracies in the address field (e.g., some state names are incorporated into street addresses and therefore attributed to the wrong state; other articles are attributed to multiple states even though the address field should be restricted to the first author of any given article). In order to corroborate our initial assessment of regional output of high-quality scientific work, we queried other databases, such as the membership of the National Academy of Sciences (Table 1). Members of the National Academy of Sciences are "elected in recognition of their distinguished and continuing achievements in original research". Although this list includes a much smaller database than does PubMed, the regional attributes have little, if any, error. The statewide distribution of National Academy of Sciences members, normalized to the total number of advanced technical degrees (MDs, PhD scientists and PhD engineers) in that state, is shown in Figure 2.

Comparison of Figures 1 and 2 suggest that the geographic patterns of the "quality" of science converge to the same loci. Similar plots (not shown) reveal a remarkable consistency in the convergence of both scientific and technological productivity when other parameters, such as the number of new utility patents, are examined. These data appear to support the hypothesis that high performance in science and technology is an outcome of attaining a "critical mass" of human and infrastructural resources. This hypothesis proposes that scientific discoveries and technological developments are unlikely to occur where human and infrastructural resources are sparse, but that the probability of scientific and technological breakthroughs occurring in a given area increases as the overall scientific and technological activities intensify. This seems intuitive, as are other correlates of this hypothesis. The presence of several internationally prominent, well-funded researchers at a given institution will attract highly motivated students, post-doctoral candidates and junior faculty to that institution. If the science and engineering departments of an academic institution are highly attractive, and the surrounding community has a highly desirable quality of life, then there will be an accumulated labor pool of highly talented and well-educated individuals, which in itself becomes an attractive feature for seed capital investment and for start-up technology firms seeking a specialized workforce.

To assess this hypothesis more critically, we analyzed the raw data in Table 1 using the Spearman rank order correlation. In this analysis, paired variables with positive correlation coefficients and P values below 0.050 are related and tend to increase together; P values greater than 0.050 are not statistically significant. A "perfect" correlation would have a Spearman correlation coefficient ([r.sub.S]) of 1.000 and a P value of 0.000. Assume that the number of "high-profile" scientific publications and the number of members in the National Academy of Sciences reflects the quality of science, and that the number of total scientific publications, scientific PhDs, engineering PhDs, and non-federal MDs reflect the quantity of science. Which other parameters, such as population, economic output, per capita income, etc., are most closely related to a high quality of science (9)?

The absolute number of "high-profile" scientific manuscripts is most closely correlated with the total number of scientific manuscripts ([r.sub.S] = 0.922, P = 0.000). This is also closely associated with the number of new NCI research grants ([r.sub.S] = 0.908, P = 0.000) and the number of NAS members and the number of PhD scientists (both have [r.sub.S] = 0.902, P = 0.000). Interestingly, there is a much stronger correlation between the absolute number of "high-profile" scientific manuscripts with either the total R & D performance ([r.sub.S] = 0.888, P = 0.000) or by the gross state product ([r.sub.S] = 0.861, P = 0.000), than with the normalized ratio of these economic parameters, referred to as "research intensity" ([r.sub.S] = 0.579, P = 0.000).

If one normalizes the "high-profile" scientific publications to the total number of publications, then one observes the strongest correlation with the highest "quality" of highly trained technical personnel (defined here as the percentage of NAS members relative to the total number of scientific PhD, engineering PhD and non-federal MDs). This correlation ([r.sub.S] = 0.731, P = 0.000) is surprisingly much higher than that observed between total R & D performance ([r.sub.S] = 0.486, P = 0.000), gross state product ([r.sub.S] = 0.382, P = 0.00592), or with research intensity ([r.sub.S] = 0.426, P = 0.00192).

What does this mean? The simplistic, and perhaps correct, interpretation is that high quality scientific productivity is more dependent on a pool of highly talented individuals than upon research funding alone. This provides a modicum of hope for a poor state such as Mississippi, because it provides insight into how we can improve our scientific standing. It is easy to illustrate the fundamental concept using the team sports analogy. Every athletic director knows that the quality of a school's football program is not determined by the amount of money spent on helmets, jerseys and transportation to travel to away games. Granted, there is a critical spending threshold that must be exceeded if one wants to have a football team, but the key ingredients are the athletes. Good athletes, not good helmets, win games and define themselves as good teams. One great athlete does not constitute a great team. Excellent teams find it easier to recruit excellent athletes. These truisms are accepted throughout the Southeastern Conference and other collegiate conferences in the south. Yet despite their obvious nature, the data suggest that these concepts have not been applied towards building strong scientific programs in the southeastern United States.

At some point, the "critical mass" hypothesis becomes self-fulfilling and begins to influence private policy decisions. For example, there are several private research funding agencies that fund researchers from a list of invited institutions. "The Searle Scholars Program makes grants to selected universities and research centers to support the independent research of exceptional young faculty in the biomedical sciences and chemistry" (quoted from the Searle Scholars website, http://www.searlescholars.net/index.html). None of the 123 institutions invited to nominate Searle Scholars is located in Mississippi (http://www.searlescholars.net/apply/participating_inst.html). Furthermore, the density of Searle Scholar awardees fits neatly into the recurring geographic distribution patterns of other scientific performance parameters, such as National Academy of Science memberships, as shown in Figure 3.

This pattern is repeated in the Pew Scholars Program in the Biomedical Sciences (http://futurehealth.ucsf.edu/pewscholar.html), which "is designed to support young investigators of outstanding promise in the basic and clinical sciences relevant to the advancement of human health." None of the 132 institutions invited to nominate investigators for the Pew Program is located in Mississippi (http://futurehealth.ucsf.edu/biomed/institutions.html). More significant than the distribution of invited institutions is the geographic distribution of the Pew Scholars (Figure 3), which is strikingly similar to that of the Searle Scholars. If the private sector that voluntarily rewards scientific innovators tends to avoid Mississippi entirely, how can one realistically expect the private sector that invests in technological innovators to be attracted to Mississippi?

We began the first half of this article with an attempt to objectively measure the collective scientific performance of Mississippi. In the course of this effort, we uncovered evidence that a non-uniform geographic distribution of scientific activity arises from a self-reinforcing mechanism which benefits "high-quality" activities and penalizes "low-quality" activities. In the remainder of this article, we will try to address how Mississippi scientists and Mississippi policy makers can exploit this hypothesis to set a course through the uncharted future of biomedicine.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

THE NATIONAL CENTER FOR NATURAL PRODUCTS RESEARCH

The dilemma facing Mississippi's biomedical research community is a classic "catch-22" problem: how can we improve research productivity, and thereby attract more research infrastructural resources, if more resources are required to improve research productivity? Obviously, we start with what we have. One of Mississippi's unique resources is the National Center for Natural Products Research (NCNPR), based at the School of Pharmacy at the University of Mississippi's Oxford campus. The NCNPR is housed in the Thad Cochran Research Center, a state-of-the-art facility that provides pivotal infrastructure for a school-wide research enterprise. Currently, approximately 120 full-time research personnel are employed, including 35 federal employees of the U.S. Department of Agriculture's Agricultural Research Service (USDA/ARS). Also integral to the Center's research programs are 25 academic faculty of the School of Pharmacy, who have joint appointments in NCNPR. For the purpose of our article, the NCNPR brings two important perspectives: (1) the history of the NCNPR is instructive in terms of implications for building new capacity and infrastructure and improving research productivity; and (2) with regard to the research program, it can serve an important role in the growth of related science and technology in Mississippi.

The development of the NCNPR illustrates how an essential combination of vision, leadership, state and university support, a creative scientific community, long-term commitment, and an investment in infrastructure can lead to successful scientific and technological enterprises.

In the 1960s Charles Hartman, Dean of the School of Pharmacy at the university, had a vision for Mississippi that incorporated the state's pharmacy research leadership with its agricultural economy and resources. He carried his idea to the state legislature, and in 1962 the Research Institute of Pharmaceutical Sciences (RIPS) was created. It was a small step and a modest investment, and its charter called for an institute that would discover, develop and bring to the market pharmaceuticals based on natural products. Over the 1970s and 1980s, the RIPS program grew steadily. Another significant development was the decision to maintain Pharmacognosy as a distinct program at the School of Pharmacy. Pharmacognosy is the science that deals with drugs derived from natural sources, including plant, animal and microbial sources. During a time when many schools were eliminating or merging their Pharmacognosy departments, and removing much of this content from their curricula, Ole Miss not only maintained, but strengthened the program, recruiting research leaders to the department. The University of Mississippi is one of only twenty institutions in the United States that offers a graduate degree in Pharmacognosy or Natural Products Chemistry. The latter half of the 20th century was 'the age of chemistry,' and the focus in new pharmaceutical discovery shifted considerably away from crude botanical drugs and more toward synthetic chemistry; it was widely regarded that the cure for most diseases would be realized in the laboratory design and synthesis of new 'magic bullets.' When many universities were closing or merging their pharmacognosy programs, the University of Mississippi continued a major emphasis in pharmacognosy, developing a world-wide reputation in this discipline.

For more than 25 years, these sustained efforts to build a research enterprise, recruit and retain research leaders, and maintain the educational focus in pharmacognosy, have required extraordinary vision, leadership, persistence, and a significant investment in research infrastructure. The commitments from the state, the university, and the school were critical investments, and ultimately afforded an opportunity, in partnership with the U.S. Department of Agriculture, to establish the National Center for Natural Products Research.

Research activities at the NCNPR cover a broad range of activities that include a natural products discovery and development program, with both pharmaceutical and agrochemical applications. In addition, a medicinal plant research program is aimed at studying botanical, agronomic, chemical, and pharmacological aspects of plants that are sources of pharmaceuticals, or of botanical dietary supplements. The research efforts at the NCNPR apply to medical research that is within the scope of this article, but also to non-medicinal ventures that have tremendous importance to the agricultural sector of Mississippi's economy. There are two aspects of the NCNPR that are particularly relevant to future biomedical research ventures in Mississippi. First, the NCNPR was specifically instituted with a mission to commercialize products or technologies developed by the Center. This includes the capability to protect intellectual property developed, to manage an intellectual property portfolio, to develop license agreements with the private sector, and to prepare technology dossiers that will support investigational new drug (IND) applications.

As a result of its research activities, the NCNPR has six active license agreements as of July, 2004, and four of these agreements currently generate revenue for the center. This includes license agreements for such products as Immulina, an algal-derived immunostimulant formulation, and dronabinol hemisuccinate, a plant-derived anti-emetic which can be administered orally or in suppository form to prevent nausea and vomiting in cancer patients. Eleven additional patent applications are pending. Several other products are in pre-approval or pre-clinical development phases. These include an 8-aminoquinoline derivative under pre-clinical evaluation for the treatment of malaria, leishmania and pneumocystis pneumonia. Also, a natural product-derived algaecide is undergoing testing in catfish ponds to prevent blue-green algae from adversely affecting the flavor of catfish; catfish is Mississippi's fourth largest agricultural commodity. The comparable capability to translate basic research into clinically useful and potentially marketable technologies will be extremely important for the University of Mississippi Medical Center Cancer Institute (UMMCCI) to develop, as discussed below.

The second aspect relevant to future biomedical research ventures is that the NCNPR has accumulated both capital-intensive core research instrumentation and talented personnel required to productively exploit this instrumentation. The NCNPR has a central repository containing over 20,000 natural product samples, including a variety of extracts and pure isolates. These materials can be manipulated for high-throughput screening assays and related purposes with robotic workstations that integrate archival retrieval, sample processing and assay readout functions. The central instrumentation facilities, which serve both the Center and the School of Pharmacy, are equipped with nuclear magnetic resonance spectrometers with operating frequencies of 400, 500, and 600 MHz. These facilities also have a Fourier-transform ion cyclotron resonance (FT-ICR) high resolution mass spectrometer, liquid chromatography mass spectrometers, and a full complement of additional chromatographic and structure elucidation instruments. The biological screening facilities permit cellular, biochemical, and microbiological assays that can be conducted in a relatively high throughput fashion. The greenhouses, experimental beds, field plots, and analytical labs afford infrastructure for detailed agronomic or horticultural studies of medicinal plants. Finally, facilities for scaling up extraction and isolation procedures allow developmental studies for commercial applications. In short, the NCNPR and the associated departments in the School of Pharmacy have virtually any technological component required in the study of the chemistry and biology of natural products, and are staffed by a cohort of creative and highly motivated faculty and scientists in pharmacognosy, medicinal chemistry, pharmacology, and pharmaceutics. The NCNPR is an excellent example of how one can intentionally design a research facility in Mississippi to further build upon an area of existing expertise. This model began with a preexisting pool of talented researchers with an established reputation for excellence, and then added capital resources to allow these researchers to continue conducting the highest quality research possible. While state funding for the NCNPR is still at the 1995 level, the operating budget has increased approximately four-fold, and the staff has increased from 24 employees in 1995 to 85 employees today. Furthermore, two private businesses are now operating in Oxford largely because of the NCNPR's attractive "critical mass" quality, and several other biotechnology, pharmaceutical and start-up companies are seriously contemplating a decision to locate their businesses in Oxford. Thus, this research center has a positive impact on the local economy, both directly and indirectly. So now we move on to the next question: How do we build a nationally competitive research program in an area in which Mississippi has limited expertise?

THE UNIVERSITY OF MISSISSIPPI MEDICAL CENTER CANCER INSTITUTE

The University of Mississippi Medical Center Cancer Institute (UMMCCI) is a bold new initiative which proposes to remedy the statewide problem of cancer by forging new therapies from scientific discoveries yet to be made. It is estimated that 15,120 Mississippians will be newly diagnosed with cancer (excluding basal and squamous cell skin cancers and non-bladder carcinomas in situ) in 2004, and that 6,230 people in our state will die from cancer in 2004 (American Cancer Society, 2004). The total economic burden of cancer on the nation was $143.5 billion in 1996 (National Cancer Institute, 2004). Cancer provides the best example of how science and technology converge in biomedicine, because so many scientific and engineering disciplines must be seamlessly integrated in the effort to understand, prevent, diagnose and treat cancers. The recognition that cancer is a disease rooted in genetic instability came after years of collaborative research conducted by geneticists, biophysicists, developmental biologists, analytical chemists and epidemiologists. Pharmacologists, molecular biologists, medicinal chemists, oncologists, X-ray crystallographers, and veterinarians have worked together as teams to develop novel molecularly-targeted drugs such as imatinib (Gleevec[R]), the first small molecule targeted protein-tyrosine kinase inhibitor to be used in cancer chemotherapy (Smith et al., 2004). The development of advanced technology used to diagnose and treat cancer patients, such as positron emission tomography (PET), required the work of nuclear physicists, materials scientists, electrical engineers, computer scientists and mathematicians in cooperation with biological scientists. The "high technology" of physics will continue to improve cancer therapy, as illustrated by new technologies being developed at the UMMC. These include computer tomography (CT)-guided radiofrequency ablation of lung tumors (Steinke et al., 2004) and magnetic resonance image-guided cryoablation of renal tumors (Sewell et al., 2003).

The UMMCCI should not only become the state's preeminent site of cancer-related science and technology development, but a competitive national force as well. If Mississippians can identify the critical scientific questions that have not been successfully answered, and develop new and broadly affordable diagnostic and therapeutic technology, then our state can lead the field of oncology in the 21st century. Here are a few of the challenges that face Mississippi's cancer research community.

      (1) How can we effectively and affordably diagnose all cancers
      at their earliest identifiable stage?
      (2) How do we predict the therapeutic regimen that will most
      likely lead to a successful outcome for any given patient?
      (3) How can we overcome therapeutic (i.e., chemotherapy or
      radiation) resistance mechanisms?
      (4) How can we prevent metastasis?
      (5) How can we prevent cancers?

A few points should be considered to put these questions into the proper perspective. Each unique human genetic script is written in 3 billion DNA base pairs on 46 chromosomes (22 pairs of autosomal chromosomes and 2 sex chromosomes). As a fertilized human egg divides and differentiates into the approximate 50 trillion cells that comprise the adult human body, the genetic script is selectively expressed in these individual cells. Accordingly, the identity of each cell is largely defined by a network of approximately 30,000 to 40,000 proteins which must function together in a well-coordinated fashion. Given the magnitude of events that must be carefully orchestrated for a human to properly develop and live healthfully, there are numerous possible molecular defects that can serve as entry points for the development of cancer (Kinzler and Vogelstein, 1996; Hanahan and Weinberg, 2000).

With such a dizzying array of possible molecular causes of cancer, how does one begin to address the five major challenges listed above? One unifying answer may be found in the area of molecular diagnostics, which relates to the field of functional genomics. Gene microarray technology (Schena, et al., 1995) is based on the simple and well-established principle of complementary DNA strand hybridization (Perry et al., 1964). The application of multicolor fluorescent labels, together with improvements in microfluidics, high-speed robotics, high spatial resolution image analysis, and the availability of a well-annotated human genome have combined to make it possible to determine simultaneously the relative expression pattern of over 19,000 characterized and unknown human ESTs (expressed sequence tags). One of the immediate applications of this technology was to determine the changes in gene expression upon the experimental suppression of tumorigenicity in a human melanoma cell line (DeRisi et al., 1996). Since then, approximately 37% of the nearly 5,000 manuscripts discussing gene microarray experiments have been related to cancer research.

This genomic information can be further enriched by combining it with data from proteomic experiments (Liotta and Petricoin, 2000). In contrast to the well-established use of complementary DNA hybridization as the fundamental consensus technology for genomics, the technology of choice for proteomics is still evolving. The future of proteomics may involve immobilized microchip arrays of monoclonal antibodies or polynucleotide aptamers as sensitive and selective biosensors (Brody et al., 1999), or it may involve the analysis of mass spectroscopy ion signatures (Geho et al., 2004; Schwartz et al., 2004). Protein profiles taken from peripheral blood samples may soon provide new ways to detect early-stage cancers. The identification of novel tumor markers from proteomic studies may lead to the development of novel monoclonal antibodies for immunohistochemical confirmation of suspected tumor biopsies or for cancer therapy. Another likely outcome of proteomic studies of early stage cancers will be the development of marker-specific radiotracers for positron emission tomography (PET) imaging (Collier et al., 2002) to locate the tumor and assess the extent of disease.

In the opinion of the authors, Mississippi has some of the most dedicated and talented medical personnel that can be found anywhere in the nation. However, because Mississippi has so few basic scientists involved in well-funded cancer research, the model for building a successful basic research program at the UMMCCI cannot simply follow the model of the NCNPR. The recruitment and retention of human resources, in addition to the acquisition of technological hardware, is vital to the success of the UMMCCI. A key milestone will be to increase substantially the amount of peer-reviewed research activities funded by the National Cancer Institute. The statewide distribution of new NCI grants is presented in Figure 4, and there is a tremendous chasm between the low (one grant) and high (752 grants) values. Spearman correlation analysis of the data in Table 1 provides an interesting twist to the patterns observed thus far. The number of new NCI grants is most strongly correlated to the total number of PubMed citations ([r.sub.S] = 0.942, P = 0.000) and to the number of PhD scientists ([r.sub.S] = 0.927, P = 0.000). The correlations to number of "high profile" publications ([r.sub.S] = 0.908, P = 0.000) and current NAS members ([r.sub.S] = 0.857, P = 0.000) are both marginally weaker. One must exercise caution when interpreting these results. For example, highly respected specialized cancer research journals might provide a stronger Spearman correlation value than the "high profile" journals we selected. Similarly, the number of National Academy of Sciences members with oncology-related scientific disciplines might have provided a stronger correlation value than the analysis based on all NAS members. With such caveats in mind, it appears that the amount of scientific effort might be the most important factor in the procurement of peer-reviewed NCI grants.

This observation does not imply that shoddy efforts will receive peer-reviewed funding. It does suggest that we can pursue the goal of establishing the UMMCCI's basic science program concurrently with our efforts to raise the caliber of scientific activity within our state. It is important to remember that the "quality" of human endeavors is not a fixed determinant, but that "quality" is malleable and can change over time. For this reason, researchers who have a reputation of producing high quality research can never rest on their laurels, and less-acclaimed researchers can dramatically improve their reputations by designing clever experiments which yield novel discoveries.

The basic research program of the UMMCCI will require a substantial investment of capital equipment and facilities in order to provide the technological resources to conduct competitive research. Whole animal imaging systems, laser capture microdissection tools, high-throughput DNA sequencing capability, and cell and tissue repositories will be needed by cancer researchers regardless of their individual specialties. Yet the most important component of building the basic research program will be an aggressive, multi-level personnel development effort. The UMMC has a relatively small, but exceptionally talented pool of clinicians who provide a pre-existent collection of potential collaborators for translational research projects. In order for the translational bridge to be completed, however, the pillars of basic cancer research must be equally well-grounded. While Mississippi has many excellent basic scientists, many have left the state for more lucrative positions elsewhere. Those expatriate scientists often left with an intact and active research program in tow, so the "brain drain" must be corrected or we will continue to lose some of our most talented individuals. Moreover, new investigators must be given an adequate start-up package, including equipment, support staff, "protected time" for research and a reasonable budget for materials and supplies that will provide them with a few years of productive research to become competitive researchers. Finally, we must recruit a small cluster of nationally recognized, high-quality researchers. Such individuals will not only be essential to rapidly building a well-funded basic science program, but they will also serve as magnets to attract other researchers at the post-doctoral and graduate student level. The late Arthur C. Guyton, an exceptionally distinguished researcher and educator who was once president of the Mississippi Academy of Sciences, was such a magnet in the Department of Physiology and Biophysics at the University of Mississippi Medical Center (Hall et al., 2003). The presence of such individuals will make it easier to recruit other nationally competitive researchers, because they will stand as proof of Mississippi's commitment to high quality research, and they will also serve as the epicenter of the UMMCCI's critical scientific mass.

CONCLUDING REMARKS

It is incumbent upon the scientific community (i.e., the membership of the Mississippi Academy of Sciences) to raise public awareness as to why it is in the best interests of the State to support high quality scientific ventures. As scientists, we are obligated to present our arguments truthfully and without distorting the facts to curry support. We must do a better job of raising the level of scientific literacy in Mississippi to improve statewide public support for science, and we must redouble our efforts to attract private sector support. But perhaps most importantly, we should not forget that we, the scientific community, are the most important ingredient in any recipe for the advancement of science in Mississippi. When planning for Mississippi's future in biomedicine, we should remember that the finest incubator in the world won't hatch any chickens if we forget the eggs. This is as true for science as it is for poultry farming.

Table 1. Selected indicators of scientific activity in the U.S.A. These
data were derived from the following sources as follows. PubMed
citations from the inclusive period of 1990 through 2000 were attributed
to each state via the affiliation search field tag [ad] and "high
profile" citations were further restricted to the journal titles listed
in the text via the journal title search field tag [ta]. The statewide
distribution of National Academy of Science members was obtained from
the NAS website (http://www4.nationalacademies.org/nas/naspub.nsf/
urllinks/NASLocation?OpenDocument&count=500000) on March 18, 2004. New
NCI research grant data from the inclusive period of 2000 through 2003
were collected from the CRISP database (http://crisp.cit.nih.gov/crisp/
crisp_query.generate_screen) with the limitations of "new" in the "award
type" field and "research grants" in the "activity" field. The statewide
census data on PhD scientists (2001), PhD engineers (2001), population
(2002), gross state product (2000), per capita income (2001) and utility
patents were derived from the "Science and Engineering State Profiles:
2000-2001" (http://www.nsf.gov/sbe/srs/nsf03324/tables/table1.xls)
compiled by the National Science Foundation's Division of Science
Resources Statistics. The statewide census of nonfederal physicians was
derived from the Henry J. Kaiser Family Foundation's State Health Facts
Online database (http://www.statehealthfacts.org).

               "High profile"                    Current
                  PubMed          PubMed           NAS         2002
                 citations       citations       members     population
STATE           (1990-2000)     (1990-2000)    (March 2004)  (thousands)

Alabama             161            13398             2           4487
Alaska               12              682             0            644
Arizona             239            11762            26           5456
Arkansas             23             5084             0           2710
California         5731           124274           579          35116
Colorado            414            16445            30           4507
Connecticut         370            17516            72           3461
Delaware             20             2904             4            807
Florida             208            25902            24          16713
Georgia             163            19984            10           8560
Hawaii               60             2997             5           1245
Idaho                 8              972             1           1341
Illinois            482            34728            82          12601
Indiana             163            16083            13           6159
Iowa                172            16768             8           2937
Kansas               63            10486             2           2716
Kentucky             52             9106             0           4093
Louisiana            37            11146             4           4483
Maine                23             1714             6           1294
Maryland            726            52221            90           5458
Minnesota           281            26036            12           5020
Mississippi           7             4546             0           2872
Missouri            193            19851            22           5673
Montana              22             1438             1            909
Nebraska             52             7344             3           1729
Nevada               32             1482            10           2173
New Hampshire        26             2410            10           1275
New Jersey          359            15981           110           8590
New Mexico           55             5030            10           1855
New York           2960           102498           198          19158
North Carolina      478            33023            38           8320
North Dakota          2             1189             0            634
Ohio                170            35213            12          11421
Oklahoma             70             6952             1           3494
Oregon              341            11481            15           3522
Pennsylvania        794            51443            74          12335
Rhode Island         34             4673             9           1070
South Carolina       44             7867             2           4107
South Dakota          0              794             0            761
Tennessee           165            16544             9           5797
Texas              1150            66912            46          21780
Utah                235             9843             7           2316
Vermont              51             3734             1            617
Virginia            205            24939            14           7294
Washington, DC      158            16929            16            571
Washington         1372            48196            58           6069
West Virginia         5             3189             0           1802
Wisconsin           382            23761            41           5441
Wyoming               5              791             1            499

                 2001       2000          2000
                 per     total R & D   gross state    2001      2001
                capita   performance     product     utility     PhD
STATE           income   (billion $)   (billion $)   patents  scientists

Alabama          24589      1.73           120          382      5040
Alaska           30936      0.196           28           50      1350
Arizona          25872      3.107          156         1540      6720
Arkansas         22887      0.454           68          180      2670
California       32702     55.093         1345        18598     70650
Colorado         33470      4.23           168         1927     12150
Connecticut      42435      4.888          159         1853      9620
Delaware         32472      1.532           36          382      3530
Florida          28947      4.663          472         2649     16330
Georgia          28733      2.796          296         1370     11860
Hawaii           29002      0.291           42           95      2550
Idaho            24621      1.434           37         1697      2090
Illinois         33023     12.767          467         3640     20680
Indiana          27783      3.252          192         1358      9080
Iowa             27331      1.017           90          751      4500
Kansas           28565      1.42            85          312      4170
Kentucky         24923      0.866          119          481      4950
Louisiana        24535      0.627          138          520      5270
Maine            26723      0.319           36          145      2120
Maryland         35188      8.634          186         1483     22150
Minnesota        33101      4.299          185         2635     10680
Mississippi      21750      0.513           67          166      2930
Missouri         28226      2.583          179          841      8850
Montana          23963      0.17            22          145      1730
Nebraska         28886      0.439           56          215      2820
Nevada           29897      0.377           75          313      1790
New Hampshire    34138      0.775           48          598      2350
New Jersey       38506     13.133          363         3869     20660
New Mexico       23155      3.085           54          376      6800
New York         36019     13.556          799         6349     42610
North Carolina   27514      5.045          282         1946     16780
North Dakota     25902      0.146           18           97      1150
Ohio             28816      7.662          373         3274     18580
Oklahoma         25071      0.66            92          576      4240
Oregon           28165      2.116          119         1259      7260
Pennsylvania     30720      9.842          404         3534     24630
Rhode Island     30215      1.501           36          287      2370
South Carolina   24886      1.126          113          565      5030
South Dakota     26664      0.085           23           76      1160
Tennessee        26988      2.057          178          813      8680
Texas            28581     11.552          742         6371     28610
Utah             24180      1.361           69          715      4700
Vermont          28594      0.465           18          453      1800
Virginia         32431      5.069          261         1115     16960
Washington, DC   40150      2.296           59           67     13410
Washington       32025     10.516          220         1969     14540
West Virginia    22881      0.457           42          148      1980
Wisconsin        29270      2.693          173         1837      8520
Wyoming          29416      0.061           19           51       940

                                                    New NCI
                     2001            2002          research
                      PhD         Nonfederal        grants
STATE              engineers      physicians      (2000-2003)

Alabama              1340            9365              89
Alaska                 80            1341               3
Arizona              2000           11912              79
Arkansas              370            5468              16
California          21040           90671             752
Colorado             2070           11357             121
Connecticut          1410           12742              63
Delaware              840            2173               7
Florida              3080           42892              88
Georgia              1780           18403              61
Hawaii                310            3609              22
Idaho                 570            2305               1
Illinois             3940           35415             177
Indiana              1790           13463              52
Iowa                  560            6186              39
Kansas                550            6270              17
Kentucky              450            9232              29
Louisiana             870           11658              28
Maine                 280            3748               7
Maryland             3440           21018             245
Minnesota            1950           13540             136
Mississippi           660            5145               3
Missouri             1440           15024              87
Montana               100            2049               3
Nebraska              330            4084              21
Nevada                540            4085               5
New Hampshire         650            3350              38
New Jersey           4690           28799              85
New Mexico           2340            4214              29
New York             6490           76567             512
North Carolina       2340           20455             198
North Dakota          130            1497               2
Ohio                 4780           31841             219
Oklahoma              920            7072              14
Oregon               1460            8961              50
Pennsylvania         4650           41301             364
Rhode Island          500            3835              68
South Carolina        980            9234              42
South Dakota           90            1608               4
Tennessee            1660           14858             130
Texas                8910           47201             336
Utah                 1220            4797              36
Vermont               240            2181              11
Virginia             3400           18953             100
Washington, DC       1150            3938              86
Washington           2610           15761             224
West Virginia         380            4554               7
Wisconsin            1610           13746              71
Wyoming               100             928               3

ACKNOWLEDGMENTS

The authors wish to acknowledge the helpful suggestions of Dr. David Dzielak and Dr. Lyn Stabler. The research of R.J.D. is supported by a Research Scholar Grant (RSG-01-060-01-CDD) from the American Cancer Society.

(1) Editorial note: This is the first of an occasional series of articles jointly authored by members of the Mississippi Academy of Sciences and the Mississippi Technology Alliance to examine contemporary issues in science and technology.

(8) Manned space flights are technologically successful, although the number of rocket dealerships in any given city would hardly be indicative of this success.

(9) The Spearman rank order correlation provides only a simple pairwise analytical approach and is used here to stimulate thought and discussion; cluster analysis might provide a more meaningful understanding of how multiple interacting factors can contribute to an environment that fosters high quality science. Indeed, that very cluster building and analysis strategy is being pursued as a key component of the work of the Mississippi Technology Alliance.

LITERATURE CITED

American Cancer Society. 2004. Cancer Facts and Figures 2004. Atlanta, Ga: American Cancer Society.

Brody, E.N., M.C. Willis, J.D. Smith, S. Jayasena, D. Zichi, and L. Gold. 1999. The use of aptamers in large arrays for molecular diagnostics. Molecular Diagnostics 4:381-388.

Collier, T.L., R. Lecomte, T.J. McCarthy, S. Meikle, T.J. Ruth, F. Scopinaro, A. Signore, H. VanBrocklin, C. van De Wiele, and R.N. Waterhouse. 2002. Assessment of cancer-associated biomarkers by positron emission tomography: Advances and challenges. Disease Markers 18:211-247.

DeRisi, J., L. Penland, P.O. Brown, M.L. Bittner, P.S. Meltzer, M. Ray, Y. Chen, Y.A. Su, and J.M. Trent. 1996. Use of a cDNA microarray to analyse gene expression patterns in human cancer. Nature Genetics 14:457-460.

Galilei, Galileo. 1610. Sidereus Nuncius (Starry Messenger).

Galilei, Galileo. 1632. Dialogue Concerning the Two Chief Systems of the World--Ptolemaic and Copernican.

Geho, D.H., E.F. Petricoin, and L.A. Liotta. 2004. Blasting into the microworld of tissue proteomics: a new window on cancer. Clinical Cancer Research 10:825-827.

Hall, J.E., A. W. Cowley, Jr., V.S. Bishop, D.N. Granger, L.G. Navar, and A.E. Taylor. 2003. In memoriam. Arthur C. Guyton (1919-2003). Physiologist 46:126-128.

Hanahan, D., and R.A. Weinberg. 2000. The hallmarks of cancer. Cell 100:57-70.

Kinzler, K.W, and B. Vogelstein. 1996. Lessons from hereditary colorectal cancer. Cell 87:159-170.

Liotta, L. and E. Petricoin. 2000. Molecular profiling of human cancer. Nature Reviews Genetics 1:48-56.

Mississippi Technology Alliance. 2004. The Mississippi Innovation Index: A tool to measure, monitor and promote progress. Jackson, Mississippi.

National Cancer Institute. 2004. Cancer Progress Report--2003 Update. National Cancer Institute, NIH, DHHS, Bethesda, MD, http://progressreport.cancer.gov/.

National Science Board. 2002. Science and Engineering Indicators--2002. Arlington, VA: National Science Foundation (NSB-02-1).

National Science Foundation. 2003. Science and Engineering State Profiles: 2000-2001 (NSF 03-324) Note: this document is only available on the internet (http://www.nsf.gov/sbe/srs/nsf03324/start.htm#data).

Oeppen, J. and J.W. Vaupel. 2002. Broken limits to life expectancy. Science 296:1029-1031.

Payson, S. 1999. National Patterns of R & D Resources: 1998, Arlington, VA: National Science Foundation (NSF 99-335).

Perry, R.P., P.R. Srinivasan, and D.E. Kelley. 1964. Hybridization of rapidly labeled nuclear ribonucleic acids. Science 145:504-507.

Schena, M., D. Shalon, R.W. Davis, and P.O. Brown. 1995. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270:467-470.

Schwartz, S.A., R.J. Weil, M.D. Johnson, S.A. Toms, and R.M. Caprioli. 2004. Protein profiling in brain tumors using mass spectrometry: feasibility of a new technique for the analysis of protein expression. Clinical Cancer Research 10:981-987.

Sewell, P.E., J.C. Howard, W.B. Shingleton, and R.B. Harrison. 2003. Interventional magnetic resonance image-guided percutaneous cryoablation of renal tumors. Southern Medical Journal 96:708-710.

Smith, J.K., N.M. Mamoon, and R.J. Duhe. 2004. Emerging roles of targeted small molecule protein-tyrosine kinase inhibitors in cancer therapy. Oncology Research 14:175-225.

Steinke, K., P.E. Sewell, D. Dupuy, R. Lencioni, T. Helmberger, S.T. Kee, A.L. Jacob, D.W. Glenn, J. King, and D.L. Morris. 2004. Pulmonary radiofrequency ablation--an international study survey. Anticancer Research 24:339-343.

Roy J. Duhe (2,3), Fazlay Faruque (4), Larry A. Walker (5), Joe C. Files (6), and Andy Taggart (7)

(3) University of Mississippi Medical Center, Jackson, MS 39216; (4) Geographic Information Systems, University of Mississippi Medical Center, Jackson, MS 39216; (5) National Center for Natural Products Research, University of Mississippi, University, MS 38677; (6) University of Mississippi Medical Center Cancer Institute, Jackson, MS 39216; and (7) Mississippi Technology Alliance, Ridgeland, MS 39157

(2) Correspondence and requests for reprints to R. J. Duhe, Department of Pharmacology and Toxicology, University of Mississippi Medical Center, Jackson, MS, 39216-4505, USA. Tel. [phone]; Fax [phone]; e-mail: [e-mail]