The Poeciliopsis Heat Shock Model: The Integrative Strategies Behind the Data

By Lawrence E. Hightower
Molecular & Cell Biology Dept., University of Connecticut,
Storrs, CT, USA
(e-mail: hightower@biotek.mcb.uconn.edu)

 

Herein, I have attempted an integrative approach to the discussion of future research directions in the field of Stress Biology and the programmatic emphasis of the IUBS on Integrative Biology. I will use as a case study one of the first units organized to develop integrative approaches in the cellular stress response field, The Marine/Freshwater Biomedical Sciences Center, established at the University of Connecticut in 1987 for the study of molecular, cellular and organismal responses to environmental stress. The name of the center derived from the program at the U.S. National Institute of Environmental Health Sciences (NIEHS) that provided initial funding, a program originally developed to take advantage of exaggerated physiologies of marine animals with relevance to human disease. The potential value of using multidisciplinary approaches in biological research is being recognized, and for our field in particular, a workshop sponsored by the U.S. National Science Foundation was held in 1996 with the theme "Molecular, functional, and evolutionary approaches to stress and inducible stress responses: bridging the gaps." Several of us gathered around this table (Feder, Hightower, and Morimoto) were participants at the NSF workshop as well. The case study that follows addresses several issues raised in Recommendation 4 of the workshop report: "Educate the community of prospective multidisciplinary scientists about the philosophy, attitute, and practice of multidisciplinary research. This should occur at every level of instruction. Some critical elements in implementing this education are the availability of funding for postdoctoral [fellows] and midcareer training/retraining. A crucial but unresolved issue is how to enhance cooperation of diverse administrative units both within and among academic institutions; analysis of this issue should receive priority." (Feder and Berenbaum 1996).

Integrative research incorporates different levels of biological organization and thus crosses one or more boundaries of biological organisation. Many biological disciplines remain narrowly defined and nested in one level of biological order; and therefore, integrative research is almost always multidisciplinary. The boundaries between disciplines continue to diminish in large part due to the broad application of molecular biological methods and the broad use of evolutionary theory as an organizing framework. For several closely related disciplines, it is difficult to walk into a laboratory and decide whether it belongs to a cell biologist, molecular geneticist, biochemist or biophysicist. New words have been coined to celebrate our new-found freedom, such as functional genomics and physiological genomics. Of course, there are still philosophical barriers and misunderstandings that derive from a low frequency of meaningful conversations among colleagues in different disciplines. An example that comes to mind is an amusing exchange that occurred between an ecologist and a molecular biologist during the NSF multidisciplinary workshop. The ecologist volunteered that a major difference between ecological research and molecular biology is that the former is hypothesis driven whereas the latter is problem solving. This came as a great shock to the molecular biologist who had been under the impression for at least several years that failure to state a testable hypothesis in a grant proposal was a "kiss of death." Another example will serve to launch the discussion of the Poeciliopsis model.

R. Jack Schultz, a member of the Ecology and Evolutionary Biology Department, had studied tropical and desert fish in the genus Poeciliopsis for 25 years from many different perspectives including thermal stress. Even though we were located in the same building, it took 12 years for us to have the conversation that I am about to relate. What finally brought us into the same room was the prospect of additional research funding. During an initial roundtable discussion among potential participants in the NIEHS proposal, Jack heard me describe our work on heat shock proteins (Hsp) and thermotolerance. Several days later, he was standing in my lab asking questions: "What is the most exciting thing about heat shock genes?" I responded with my best material, "Heat shock genes are among the most highly conserved genes known and they are involved in acquired or inducible thermotolerance". He responded, "Are you telling me that there is no diversity among hsp70 genes in a species? Because if you are, I am telling you that they have little to do with thermotolerance in natural populations." He explained that he had already found considerable genetic variation among individuals of even the same species in their ability to survive heat shock and that a reasonable hypothesis would be that the genes involved in acquisition of thermotolerance ought to contain variation that affects protein function. The next question was "Had anyone bothered to look?". The answer that emerged after some thought was no, not in a way that would have tested the hypothesis. That is, almost everyone was working with either a few cultured cell lines or inbred organisms. We found only one published study indicating that some mammalian and avian species have different Hsp70 isoforms (Anderson et al., 1982).

Jack provided a thirty year reservoir of information on the ecology and genetics of these fish into which I dipped frequently. We agreed upon a basic strategy employing Jack's fish colony. We would first compare Hsp patterns of closely related Poeciliopsis species from the Sonora Desert of Northwestern Mexico. Jack had developed inbred strains of about a half dozen of these species. Then we would search for within species diversity, taking advantage of a peculiarity of Poeciliopsis gene transmission known as hybridogenesis. An example of this is a mating between a wild P. monacha female and an inbred P. lucida male. This results in fertile, intraspecific hybrid fish called P. monacha-lucida. During oogenesis in a hybrid female, the entire paternal genome is discarded during the mitotic cell divisions preceding meiosis, so the paternal genome is not present to undergo reassortment or recombination with the maternal genome. The result is that natural combinations of alleles are preserved in the monacha genomes, which are passed along hemiclonally. When a standardized P. lucida genome is added back upon mating, the result is a clonal vertebrate. Due to a quirk of sex determination, the entire brood of hybrid fish is female. Thus, wild monacha genomes can be captured from a natural population and amplified clonally in the laboratory. Phenotypic difference among different hybrid fish can be ascribed to genetic differences in the monacha genomes. We further decided to search for protein polymorphisms, rather than attempting to clone and sequence all members of the Poecilopsis Hsp70 family. The use of high resolution two-dimensional polyacrylamide electrophoresis of proteins denatured with sodium dodecyl sulfate and mercaptoethanol would allow us to follow several protein families at once. Based on studies of other organisms, we assumed that the two heat shock protein families most frequently linked to acquired thermotolerance, Hsp70 and Hsp27-30, would be encoded in multigene families in Poeciliopsis. We surveyed the heat shock protein patterns of six desert species using radioisotopic labeling of intact fish and primary liver cell cultures and silver staining of two dimensional gels (White et al., 1994). Later, we benefited from the collecting skills of several colleagues who managed to obtain a permit from the Mexican government, and we added two tropical species of Poeciliopsis from southern Mexico to our survey (Norris et al., 1995).

There were a large number of Hsp27-30. Every species had a unique isoform pattern and within P. monacha, there were several different patterns (Norris et al., 1997). We also found polymorphisms in stress inducible Hsp70 but not constitutive Hsc70 and Grp78 both between and within species, in what is still the most comprehensive published survey of diversity in Hsps. No strategy is without its complications, and for this one, additional experiments were needed to rule out posttranslational modifications as a source of isoforms. Using short radioisotopic pulse-chase protocols, we were able to rule out phosphorylation and processing of precursors into products. We also did Southern blots using a probe specific for inducible Poeciliopsis Hsp70, and these results were consistent with the existence of a small multigene family. Obviously this is not as direct as analysis of nucleotide sequence data for this gene family, and this was the major compromise that we made. Even today with PCR cloning and more rapid sequencing, it is still not easy, outside of major sequencing projects to obtain complete sequence information on all members of a multigene family. During the pre-funding phase of this project, which lasted almost two years, some projects were dropped and some investigators left the project. The core that persisted were those who found areas of common interest and who were able to form working relationships. Essential to this process is the willingness to be both student and teacher and the desire to think outside of the 'box' of one's main discipline. In this era, many investigators will not take this risk out of concern that their colleagues and their granting agency will view them as having lost 'focus.' Layered on this tendency has been a further narrowing of biological investigation by emphasis upon only a small number of model organisms. The genome projects have exacerbated this tendency to the point where many molecular biologists will only work on the 'chosen organisms.' This is likely to be but a temporary constriction point in biological research. New methods of DNA sequencing are under development which, if successful, will allow any genome of interest to be sequenced rapidly and economically (Alper, 1999). But for the present, it is still usually the case that one must chose organisms that have either a deeply understood organismal biology or a well-developed molecular biology. These prevailing conditions and attitudes are barriers to integrative research.

As in virtually all university-based research programs, graduate students and postdoctoral fellows played essential roles. Postdoctoral fellows were recruited to add expertise missing from the core group. We tended to attract fellows who were interested in broadening themselves through exposure to the multidisciplinary environment of the Center. We shared graduate students by participating in their thesis committees and by allowing them to move freely among labs in carrying out their experiments. They became the essential glue that held collaborative projects together on a daily basis. The structure of the graduate school at the University of Connecticut substantially aided us in sharing students. Here, students are accepted into the graduate school, not directly into departments. They are guided by advisory committees whose members can be drawn from different departments, and even schools, within the university. I have viewed this structure at various times as both inspired and insane, but it certainly worked to our advantage in the Center, since there were no departmental boundaries with which to contend. Numerous undergraduates seeking research experiences were also attracted to the project and several completed senior honors theses. Our students were exposed to an eclectic mix that included protein biochemistry, evolutionary biology, fish physiology and genetics. It was an exciting time.

In a sense, heat shock genes and proteins were at the crossroads of the various research projects. In fact, I had approached the prospect of collaborating with the attitude that these genes were evolutionarily so highly conserved that any system under study in the Center would have a heat shock response that my lab was prepared to study at the molecular level. Our reward would be the opportunity to place this response in the context of tissue physiology and intact animals. We often hear integrative research promoted as a way to find new and unexpected information at the interface of disciplines. I will not detail all of the small advances, but I think that there were at least three findings that fulfilled this expectation. In collaboration with J. Larry Renfro, we showed that cytoprotection of transport systems of flounder kidney epithelium, induced in vitro by a sublethal pre-stress, was due to an increased transport capacity (Brown et al., 1992). When the lethal stress was applied, as much damage occurred to transporters in cytoprotected cells as in unprotected controls, but the extra capacity provided a buffer that kept the cytoprotected cells transporting at normal levels. The molecular aspect of the study was to show that flounder heat shock proteins accumulated in cytoprotected tissue. Few studies of cytoprotection have been carried out using tissue level function as an indicator. Previously, hypotheses for thermotolerance and other forms of cytoprotection invoked protection of cellular components from damage, so here the integrative approach yielded an entirely new concept.

Second, the demonstration of polymorphisms in the Hsp70 family, even within a population of the same species, was novel and has encouraged searches for diversity in other organisms. In the context of desert fish, we learned that Hsp70 polymorphisms probably had not arisen in the Sonora Desert. This is a relatively lush subtropical desert with a variety of distinct ecological niches. Our original hypothesis had been that processes of mutation and selection had occurred after ancestral fish colonized the desert streams as part of adaptation to different environments. The finding of all of the major Hsp70 desert isoforms in two species of tropical fish changed our thinking. Now, our hypothesis is that ancestral fishes, similar to modern tropical P. gracilis, were pre-adapted to high temperatures in the tropics and had Hsp70 polymorphisms prior to migrating north along the western coast of Central America to eventually reach the desert. This hypothesis fits with the biogeography of these fishes. Was this genetic diversity useful in adaptation to desert life? We do not know yet. As a first step, we asked whether or not Hsp70 levels induced in gill tissue by pre-heating correlated with amounts of acquired thermotolerance in individuals from a natural population of P. gracilis. Our use of two-dimensional gels allowed us to identify the Hsp70 isoform pattern of each fish in the study. We found that fish carrying isoform 3, the most frequently encountered Hsp70 major isoform in desert species, showed a significant positive correlation between levels of Hsp70 and thermotolerance, whereas fish without isoform 3 showed no significant correlation (Hightower et al., 1999). The literature contains numerous examples in different organisms of correlations between Hsp70 levels and acquired thermotolerance and also examples where no correlation was found. For the third novel finding, our study reproduced both of these observations in one population of individuals of the same species. Is there something special about isoform 3? This remains for future work.

Even though the value of integrative approaches is being recognized, the tasks of assembling a group of investigators across departmental and sometimes institutional lines and of finding a receptive funding agency with which to work remain daunting. The NSF workshop report provides several suggestions for lowering these barriers, and this report should be made more broadly available on the web. And what about future directions for stress research in general? Using our experiences as a case study, we think integrative research deserves a place at the table. The prevailing views of the heat shock response and acquired thermotolerance are still based on a very narrow sampling of organisms, primarily inbred laboratory strains and dedifferentiated cultured cell lines. The latter materials should be supplemented with studies of natural populations as well as tissue level studies of differentiated function as an intermediate step in connecting the molecular biology of heat shock genes and proteins to responses in animals and plants. This will be particularly important to a more complete understanding of inducible states of protection. The evolutionarily conserved nature of several heat shock gene families has encouraged the attitude that this is a research area in which we can safely concentrate on a few model organisms and find essentially everything important. This notion underestimates the power of the natural forces that drive genetic diversification and adaptation.

References