Stephen Altschul received his Ph.D. from the Massachusetts Institute of Technology in Cambridge in 1987. He came to the National Center for Biotechnology Information (NCBI) at NLM in 1989 and is currently a mathematician in its Computational Biology Branch.
My research interests center on developing improved computer methods to compare DNA and protein sequences. The problems that I study may be divided roughly into three domains: the definition of measures that reflect biological relatedness, the development of algorithms for locating similar sequence regions, and the statistical assessment of sequence similarities. These domains are interconnected, and I try to approach each with the other in mind. For example, the choice of a biologically sensible but overly complicated measure of relatedness or similarity may entail unacceptable algorithmic and statistical complications. Alternately, a statistical understanding can sometimes point to more sensitive measures and improved algorithms.
One focus of my work has been to develop rapid and sensitive methods for searching biological-sequence databases. Collaborating with scientists at NCBI and elsewhere, I helped develop the BLAST family of database search programs which are more sensitive than their predecessors. My contribution to this project centered on the statistical description of the similarities among database searches. Optimal segment-pair scores obey an extreme value distribution, and generalizations to multiple high-scoring segment pairs and multiple scoring systems are possible. These statistical results have suggested alternative definitions of sequence similarity and led to the construction of amino acid and nucleotide substitution matrices of greater sensitivity.
I have also been interested in the many questions that spring from the global or local alignment of multiple sequences. One such question is how to deal with the biases present in any collection of related biological sequences where none of the "data points" are independent, but all are correlated to one degree or another. I have helped develop the MSA program that constructs global multiple alignments and the MACAW and Gibbs programs, which are used for local multiple alignment. These latter two programs are particularly useful in the discovery of shared sequence patterns among proteins.
Most recently, I have been working on various problems that arise in the construction of position-dependent weight matrices or profiles for the description of protein motifs. Many different methods for building such matrices have been described, but a good theory to guide their construction has only recently begun to emerge.
My postdoctoral collaborators and I use the patch-clamp technique to study the physiological properties and regulation of ion channel proteins in immortalized cells from the mammalian neuroendocrine system. On the time scale of milliseconds to hours, these membrane proteins are primary determinants of neural signalling, hormone secretion and cardiovascular contractility. By measuring the ionic current through individual channel proteins as they open and close in small, cell-free patches of membrane, we have demonstrated that both voltage-activated calcium channels and calcium-activated potassium channels are regulated reciprocally by reversible protein phosphorylation.
We have also used these channels as a sensitive molecular assay to identify two new signal transduction pathways through which inhibitory neuropeptides stimulate the serine/threonine-directed protein phosphatase, PP2A. One pathway is activated by neurotransmitters like somatostatin that stimulate archidonic acid metabolism through pertussis toxin-sensitive GTP-binding proteins. The second pathway is activated by natriuretic peptides through receptors with intrinsic guanylyl cyclase activity. Understanding these pathways may have important implications for human health disorders because PP2A has been identified as the primary target of a growing number of potent microbial toxins and xenobiotics in the environment. The recent demonstration that tyrosine-directed protein kinases regulate PP2A and that somatostatin and other neuropeptides inhibit cell proliferation, suggests that the same protein phosphatase cascade may potentially modulate both electrical excitability and cell proliferation.
Our laboratory has been studying how the interactions of environmental and genetic factors can give rise to human autoimmune disorders. We are using a multidisciplinary strategy involving epidemiologic, immunologic, genetic, and molecular biologic techniques and are focusing our efforts on idiopathic inflammatory myopathies, a group of systemic connective tissue diseases marked by chronic infiltration of muscle by activated T and B lymphocytes.
Because many diseases are actually collections of different disorders grouped together by a common feature, we have been investigating ways of dividing diseases into their minimal components to understand risk factors and pathogenesis. We have discovered that the myositis syndromes are in fact composed of many distinct disorders, some of which are characterized by unique clinical or serologic features, and tend to develop in individuals who inherit specific combinations of genes encoding immunoglobulin and human leukocyte antigen (HLA) molecules.
The acute onset of myositis and the geographic clustering and seasonal associations with the onset of disease in groups of patients who make autoantibodies directed against cytoplasmic translational components imply that environmental agents may be important in initiating myositis in some patients. In addition to conducting worldwide epidemiological studies of myositis, we are investigating the possible role of environmental exposures to certain infectious agents, drugs, dietary supplements, medical devices, and occupational and other toxins as triggers of inflammatory muscle disease in susceptible individuals. Our data suggest that the development of myositis in groups of people with some of these exposures is related to the presence of specific HLA alleles that regulate immune responses. One of our current goals is to understand the mechanisms responsible for the genetic risk factors linked to these environmentally associated autoimmune diseases.
The goal of our laboratory is to understand pattern formation in Xenopus embryos. We are concentrating specifically on two groups of transcriptional regulators -- those containing the fork head and the homeobox DNA-binding domains. We isolated the first Xenopus fork head gene (XFKH1) and showed that it is likely to be involved in axis formation. We showed that this gene belongs to a larger gene family that, like the homeobox genes, seems to be involved in aspects of pattern formation and cell differentiation.
During the past five years, we have also isolated several novel homeobox genes involved in pattern formation. We are now specifically concentrating on those involved in craniofacial development. Most recently, by using the example of two novel homeobox genes specific for anterior pituitary and retinal development, we demonstrated that ammonium chloride can induce anterior regions of the amphibian head in uncommitted ectoderm. We expect this research to provide insights into the early processes of amphibian head formation.
In addition, we have initiated similar research into pattern formation in zebra fish embryos, and we have constructed a cDNA library specific for regenerating Xenopus limbs that should be helpful in understanding amphibian limb regeneration.
My laboratory studies meiotic recombination in the yeast Saccharomyces cerevisiae. In addition to playing an important role in the meiotic chromosome pairing and disjunction events that ensure the segregation of an intact haploid genome to gametes, recombination is an important component of the cell's efforts to maintain genome integrity in the face of DNA damage. Yeast are ideal organisms in which to study the molecular mechanism of recombination because the high recombination frequencies and temporal synchrony of meiosis in yeast facilitate the study of recombination events at the DNA level. We are also interested in uncovering factors and control mechanisms that determine where and when recombination occurs during meiosis.
Our general strategy in studying the mechanism of meiotic recombination has been to identify loci that display high frequencies of recombination. To help characterize molecular events that occur in the course of meiotic recombination, we introduce mutations that create both genetically scorable markers and restriction-site polymorphisms and we use physical techniques to probe DNA structure at these loci. We have used this approach to determine when and how meiotic recombination is initiated, when parental contributions to recombinants are first stably joined by heteroduplex DNA, when intermediates are resolved to form mature recombinant products, and what the structure of those products are. We plan to continue this approach to isolate and characterize early intermediates in recombination and as a tool to help determine the gene products and enzymatic activities responsible for their formation and resolution.
Our lab is also interested in the factors that determine the frequency and location of meiotic recombination events and the relationship between meiotic recombination and homolog pairing. In yeast, the distribution of meiotic exchange events is determined primarily by the location of meiosis-induced double-strand DNA breaks (DSBs), which initiate meiotic recombination. We have demonstrated a one-to-one correspondence between DSB sites and sites that display nuclease hypersensitivity in digests of chromatin. This indicates that chromatin structure plays an important role in determining where meiotic recombination is initiated and also shows the utility of DSB site analysis as a probe of chromatin structure in vivo. Experiments in progress point toward the existence of elements that act over large regions (about 3 - 5% of a chromosome) to modulate the level of recombination in a gene and also suggest that homologous chromosomes associate before the onset of meiotic recombination.
My main focus since I began my research at Professor Hideo Hayashi's lab in Japan in 1979 has been to investigate the mechanisms of leukocyte infiltration into inflammatory reaction sites. The immigration of blood leukocytes to inflammatory reaction sites appears to be mediated by chemoattractants such as N-formyl-methionyl-leucyl-phenylalanine (FMLP); C5a, a component of serum complement; and chemotactic cytokines, also known as "chemokines," which are produced at the sites. In 1987 and 1988, my colleagues at NCI and I purified and cloned two major chemokines, neutrophil attractant protein-1 (NAP-1)/interleukin-8 and monocyte chemoattractant protein-1 (MCP-1), that, as their names indicate, attract neutrophils and monocytes, respectively.
On the basis of in vitro studies and findings by immunohistochemistry and in situ hybridization on human tissues, we speculate that NAP-1 and MCP-1 are involved in various infectious diseases and tumors with neutrophil or monocyte infiltration. This would include tumors such as malignant glioma or malignant fibrous histiocytoma, which are infiltrated by macrophages. But the role of the infiltrated macrophages, and whether they are beneficial to the host, remains controversial. My current interest is in the roles of the two factors in animal models. Although mice and rats would be a first choice for this in vivo work, NAP-1 cannot be found in these animals. Therefore, we have been pursuing other animal models while continuing to study the roles of MCP-1 in rats.
After years of steady progress toward our goal of understanding these proteins, we are excited that we have finally come to a stage where we can test the effects of a neutralizing antibody against rat MCP-1 in rat disease models. I hope that my continued research at NIH will lead us to a better understanding of the involvement of NAP-1 and MCP-1 in inflammatory diseases and tumors, possibly paving the way for new approaches to the control of leukocyte infiltration that can result in tissue destruction.