| Sign In to gain access to subscriptions and/or personal tools. |
Big Events in a Small World: the Changing Face of Oral MicrobiologyOral Microbiology Unit, Department of Oral & Dental Science, University of Bristol Dental School, Lower Maudlin Street, Bristol BS1 2LY, United Kingdom; howard.jenkinson{at}bristol.ac.uk
Key Words: Bacterial adhesion • oral streptococci • gene expression
Unlike several authors of previous "Discovery!" articles, I am not able to claim any first-hand association with the microbial revolution in dental research during 1960-1980. Jason Tanzer, Else Theilade, Ron Gibbons, and Douglas Bratthall, who have all contributed to "Discovery!", were among those who shaped that revolution, during which caries was identified as a transmissible infectious disease (Tanzer, 1995) and adhesion was proposed as a driving factor in bacterial colonization of the human host (Gibbons, 1996). My contribution to dental research is more akin to a "window of infectivity", in that I have played some part in the more recent molecular revolution that has to a large extent remodeled the face of oral microbiology. Twenty years ago, I was a young post-doc in Joel Mandelstam's laboratory at the University of Oxford, England, working on the genetics and biochemistry of endospore formation by Bacillus subtilis. I was fortunate at that time to be able to learn and develop skills in the new molecular techniques of the era. These included polyacrylamide gel electrophoresis of proteins, gene cloning, restriction endonuclease mapping, and DNA sequencing. By 1980, several perspex electrophoresis systems had become commercially available. One of these systems was purchased for the research group, and I spent many months working on methods for labeling, extracting, and separating spore coat proteins (Jenkinson et al., 1981). As long as one was good at following recipes, the electrophoresis system produced very attractive protein gels. A little later, it was possible to purchase a perspex apparatus wherein protein bands could be electrolytically imprinted from gels onto nitrocellulose membrane (Western blotting), and at about the same time we acquired a DNA sequencing gel system. By 1984, we could perform the molecular laboratory miracle. A gene could be cloned, the DNA sequenced, and the encoded polypeptide blotted from gel onto membrane, from which it was accessible to sequencing, functional, or immunological analyses. I introduce "gene to protein" technology because this is about where I came into dental research. Thousands of commercial ventures have since been spawned on the back of this and associated technologies. Their perspex, plastic, polypeptide, and nucleic acid synthetics fueled the race for discovering the molecular mechanisms of adhesion, colonization, virulence, and pathogenicity in oral micro-organisms.
Joel Mandelstam was the major influence on my early career development. He had an ability to express his experiences, concepts, and ideas with a clarity that I have not really encountered since. Joel was more or less entirely responsible for my entering the field of oral microbiology research. When I was offered a Faculty position at the University of Otago School of Dentistry, in Dunedin, New Zealand, he advised me to accept it without reservation. He later confided that he had felt some measure of culpability for packing me off to the opposite face of the Earth, but I reassured him that this particular piece of advice was one of the best that I had taken. With the contract for a "real job" arriving in the post one morning at my flat in Oxford, it was with excitement that I telephoned Sue with the news that we had the opportunity to move to New Zealand. However, it was with trepidation that I asked her to consider that, according to New Zealand law at the time, she might be deported after 6 months if we were not officially tied. Thus it was that I ventured into two marriages, the second with oral microbiology. Before embarking upon our journey to the other side of the world, and into a new world of oral microbes, we had to do some research on the people, places, history, and customs that we would encounter. I had never before had any real cause to investigate New Zealand, and there were plenty of illustrated guides in the local community library to assist me. Information about staff, facilities, and research at Otago University was more difficult to come by, and had to be requested by air mail or retrieved from citation indices. Then, there was no screen button to click, which nowadays would have displayed in front of me all I wished to know (and possibly not to know) about Dunedin and Otago. The hunt for information added to the intrigue. I am now embarrassed at my surprise in learning that New Zealand is comprised of two islands and that these lay about 1400 miles (2200 km) to the east of southern Australia. Over the years I would come to appreciate the tranquillity, but also the remoteness, of New Zealand, now more often referred to by the Maori name Aotearoa (translated as land of the long white cloud). It takes a full 26 hours of flight time to travel from London, either west via the US or east via Asia, to Dunedin in the South Island of New Zealand. After leaving Los Angeles to the southwest, the flight crosses over the Pacific Ocean and the international dateline, resulting in the loss of a day in the life. After about 12 hours, it is usually possible to get a first glimpse of land, Aotearoa, "glistening like a pearl at the bottom of the world" (attributed to Tim Finn, New Zealand lyricist). Now that I am back living in England, my birthland, I presume to have retrieved that lost day, although I have for the time being lost New Zealand. At Otago University, School of Dentistry, within the Department of Oral Biology chaired by Alister Smillie, a new Unit had been set up in 1981 under the leadership of Max Shepherd, a biochemist renowned for his work on the cell wall of Candida albicans (Shepherd et al., 1985). I understood that I had been appointed essentially to inject some genetics and molecular biology into microbiology research at Otago. Although I had met Max in Oxford, and had also met George Petersen, Chair in Biochemistry at Otago (George had just published with Sanger, in Cambridge, UK, the complete nucleotide sequence of phage lambda), I had virtually no appreciation of other staff interests and facilities available within the Faculty of Medicine. I had to establish an identity within the University, develop a new research interest related to dentistry, and, most importantly, apply for funding. I decided very shortly after I arrived at Otago in 1983 that I would work on streptococci, which after all bore at least some microscopic resemblance to Bacillus spores, and my plans were stimulated by two review articles. In the first of these, Hamada and Slade (1980) educated me about Streptococcus mutans, while, in the second, Don Clewell (1981) described how streptococci sent signals to each other, mated, and could be genetically manipulated. Clearly the application of molecular genetics would provide a new approach to study of the oral streptococci. Meanwhile, Max Shepherd encouraged me to work also with his group on developing genetic methods for studying cell wall structure and virulence in Candida albicans. At times I felt torn between the two research areas of oral bacteria and yeasts, but it was essential that we collaborated: After all, united we progress, divided we digress. Although Max and I, and Richard Cannon (who joined the Experimental Oral Biology Unit in 1987 from Cambridge, UK), went on to publish a novel genetic transformation system for C. albicans (Cannon et al., 1990), we never quite "shot with the big guns" in Candida genetics. However, we also demonstrated that cells of oral streptococci and Candida specifically bound each other, or co-aggregated. Further research into the mechanisms of co-aggregation, and of C. albicans adhesion to salivary proteins, brought our laboratory groups closer.
Although Gibbons and colleagues in the US had pioneered studies on oral bacterial adhesion in the 1970s, little was known in the early 1980s about the molecular basis of adhesion processes. There was evidence from Clark (Gainesville, Florida) and Cisar (NIDR) that surface fimbrial structures on Actinomyces naeslundii (viscosus) mediated adhesion of cells to salivary components. Gibbons also promoted the notion that the "fuzzy coat", seen on the surface of some oral streptococci by transmission electron microscopy, harbored adhesive molecules. However, there was virtually no information on the molecular composition or fine structure of the oral streptococcal cell surface. Ron Doyle (Louisville) and Bob Rosan (Philadelphia) had just demonstrated that streptococcal cell wall layers contained proteins, in addition to peptidoglycan and teichoic acids. I decided that major advances could be made in understanding streptococcal adhesion, and thus dental plaque development, by purifying and characterizing cell surface protein components, cloning their genes, and generating isogenic mutants to determine protein functions. By coming in from another field, I had an open mind on the candidate adhesion genes, but my model organism had to be genetically transformable, relatively easy to cultivate, and be efficient in adhering to host molecules. It was thus that I began studies on Streptococcus sanguis, an early colonizer of salivary pellicle, and I worked chiefly with the highly transformable strain Challis. Several of my new-found colleagues were critical of my choice of bacterium, mainly because it had been isolated from blood and not human dental plaque, but also because it was not strongly associated with dental caries. I persisted, nevertheless, and while S. sanguis (strain Challis) survived a name change in 1989 to S. gordonii, I was reassured, since S. gordonii became one of the more frequently utilized oral streptococcal species in laboratory studies. In fact, S. gordonii is found at multiple sites in the oral cavity, causes infective endocarditis, and very recently has been shown to be cariogenic in rats (Tanzer et al., 2001). It is an excellent host for expression of foreign proteins, with potential therefore as a mucosal vaccine vector, and the S. gordonii Challis genome sequence is currently being completed in Australia. My first encounter with Ken Knox was at the 1985 Australian and New Zealand Divsion Meeting of IADR in Brisbane, Australia. Ken had been Director of the Institute for Dental Research in Sydney, Australia, since 1975, and he was very interested in my preliminary work on the surface properties of S. sanguis mutants, because his group had been investigating protein secretion by oral streptococci. Our meeting signaled the start of a close relationship between the Experimental Oral Biology Unit at Otago and the IDR in Sydney. Although Ken had published widely on a range of topics related to oral microbial physiology, growth, and immunology, it is his work with Tony Wicken (University of New South Wales) on lipoteichoic acids that received highest recognition (Wicken and Knox, 1975). Tony and Ken were funded in 1975 for 9 years by a grant from the US National Institute of Dental Research, which was a considerable achievement. They then went on to receive program grant funding for another 8 years from the Australian National Health and Medical Research Council–the only program in dental research in Australia. Ken frequently undertook world trips as an ambassador for oral microbiology in Australia. I looked upon him as a statesman who was somehow able to cross the scientific territories of the US, UK, and continental Europe with an open passport. His visits to leading laboratories in the US and UK were reciprocated with invitations to Sydney, and during the 1980s he hosted, among others, Gerald Shockman, Bob Fitzgerald, Bob Rosan, Arnie Bleiweis, Pauline Handley, and Phil Marsh. It was through Ken that I first met some of these big names, and, as our research in New Zealand began to be recognized internationally, so we began to interest our own visitors. Mark Herzberg, Rich Lamont, Paul Kolenbrander, and Paula Fives-Taylor each came to work in Dunedin—their visits were all big events in our small world! Although Otago and Sydney forged strong links in oral microbiology research, it was disappointing that we could not put formal mechanisms in place for joint research funding between New Zealand and Australia. In fact, it is often not appreciated, outside the region, that the two closest neighbors in the Pacific are quite far apart politically, socially, and economically.
In 1985, when our respective laboratories "down under" were working on streptococcal surface protein secretion and function, there was much publicity on the phenomenon of microbial cell surface hydrophobicity (CSH). This concept was revolutionized by Mel Rosenberg's description of a simple partitioning assay to measure bacterial adhesion to hydrocarbons (Rosenberg et al., 1980). Rosenberg and Doyle were strong proponents of the notion that CSH was related to the production of cell-surface fimbriae, and to adhesion. Enthusiasm for hydrophobicity was quite infectious, and several reports appeared that correlated CSH measurements for oral streptococci with their adhesion levels to experimental salivary pellicles. A two-stage model for adhesion was advanced, taking into account evidence for both physico-chemical (ionic and hydrophobic) interactions and the more stereo-specific interactions of protein adhesins with their receptors. However, the idea that CSH might determine adhesion was not universally popular. Some strains of streptococci (and other oral bacteria) had low levels of CSH but adhered with high affinity to salivary pellicle. CSH was influenced by growth conditions, by growth medium components adsorbed to the cell surface, and by the extent of (hydrophilic) exopolysaccharide production. Although CSH research became unfashionable, CSH provided a phenotype for cell surface protein expression. We therefore exploited hydrophobicity to identify streptococcal proteins that were potential adhesins or colonization factors. We purified from the S. sanguis cell surface a small-molecular-mass protein that partitioned into non-ionic detergents. Our excitement that this might be a novel adhesin turned into some dismay when the N-terminal amino acid sequence of this protein was found to match closely that of HPr (phosphocarrier protein) from Staphylococcus aureus. HPrs are conserved across a wide range of bacteria and are small (6-9 kDa) cytoplasmic proteins that function within the phosphotransferase system (PTS) for sugar uptake. It turned out that HPr was present within the cell cytoplasm and on the cell surface of streptococci (Jenkinson, 1989). In S. mutans, HPr corresponds to Antigen D, one of four major cell surface proteins first purified from the cell envelope of this organism by Roy Russell (Russell et al., 1983). The significance of HPr, and of other cytoplasmic proteins such as glyceraldehyde-3-phosphate dehydrogenase, being cell-surface-located is still not fully understood. It is possible that streptococci have a mechanism for extra-cytoplasmic phospho-transfer and energy transduction.
Our studies of CSH-related proteins led naturally on to an interest in determining the composition of surface structures in streptococci. Handley et al. (1985) had published arguably some of the best ever electron microscopic images of oral streptococcal fimbriae and fibrils. To understand more about the techniques involved in these studies, I spent a period of my sabbatical leave in 1989 working in Pauline Handley's laboratory in Manchester, UK. In addition to acquiring some rudimentary electronic microscopic skills, I came to appreciate that high levels of expertise, patience, and effort were necessary to produce quality micrographs! Upon my return to New Zealand in 1990, Rod McNab joined me as a post-doc from Aberdeen University, Scotland, and we began a program of identifying genes necessary for surface structure formation in S. gordonii. During 1989, there were two significant papers published that launched a decade of molecular and genetic studies on the structure and function of streptococcal antigen I/II family proteins (Jenkinson and Demuth, 1997). The major surface protein of S. mutans, designated antigen I/II by Zanders and Lehner (1981), had been the focus of anti-caries vaccine research for some years, and antibodies to this protein were protective in animal models of caries. In 1989, the groups led by Bleiweis (Gainesville, Florida) and Lehner (Guy's Hospital, London) published a joint paper describing the complete sequence of the gene encoding S. mutans antigen I/II (also designated antigen B, SpaP, P1 or Pac) (Kelly et al., 1989). In another publication that year, Don Demuth (Philadelphia) described the cloning and expression of an orthologous antigen I/II gene (designated ssp-5) from S. gordonii in Enterococcus faecalis (Demuth et al., 1989). The product of the ssp-5 (sspB) gene imparted to enterococcal cells the ability to bind salivary agglutinin glycoprotein, confirming that this was a receptor for antigen I/II family polypeptides. The combined effect of these articles was to signify that the way had been opened for more detailed analyses of streptococcal adhesin expression, structure, and function. Although spaP was the first oral streptococcal adhesin gene to be sequenced, it was not the first secreted protein gene to be sequenced. Two years earlier, Ferretti et al (1987) had published the complete sequence of a glucosyltransferase gene from Streptococcus sobrinus. In New Zealand, by 1991, we had purified another surface protein that we thought was associated with CSH in S. gordonii. We generated antibodies to this protein, cloned a portion of the gene encoding it, and then disrupted the coding sequence by insertion of an antibiotic resistance cassette (allelic replacement). The mutant cells were very much reduced in CSH, but it took several further experiments to assign this phenotype to the loss of an especially large, and difficult to extract, surface protein that we designated CshA. Subsequently, the complete cshA gene was isolated, a second related gene (designated cshB) was discovered, and various isogenic mutants in these genes were generated. We showed that CshA mediated adhesion to fibronectin and to other oral micro-organisms, and that it was essential for S. gordonii colonization of the mouse oral cavity (McNab et al., 1994). It took us a further 4 years to demonstrate unequivocally that the cshA gene contained all the necessary information to enable the 259-kDa CshA polypeptide to assemble into fibrillar structures. When cshA was expressed in E. faecalis, surface fibrils appeared on the enterococcal cells that were identical in morphology to those on the surface of S. gordonii Challis. Moreover, CshA conferred upon enterococci the properties of CSH and fibronectin binding (McNab et al., 1999). Thus, the perceptions of Doyle, Gibbons, and Rosenberg were borne out by our molecular studies demonstrating that, at least for CshA, fibril formation and CSH were co-determined. It also appeared that the CSH skeptics were vindicated, because the ability of S. gordonii cells to bind experimental salivary pellicle was not significantly affected by the loss of CshA.
Oral microbiology research in the early 1990s into adhesion, colonization, and virulence mechanisms began to be DNA sequence-driven. This is not to detract from many innovative and highly significant findings, especially in periodontal microbiology research, that did not directly involve sequencing or sequence gazing. For example, the first genetic transformation systems were established for Actinobacillus actinomycetemcomitans and Porphyromonas gingivalis. Also, it became evident that these bacteria were internalized by, and could multiply within, human epithelial cells. But recombinant gene expression technology was paramount in resolving a previous decade of questions about surface structure composition, adhesin functions, and enzyme specificities in a wide range of oral bacterial, not least in P. gingivalis! After the first few DNA sequences of surface protein genes in streptococci and staphylococci had been obtained, it became evident that the deduced amino acid sequences of the proteins showed some similar patterns and common signatures. A comparative analysis of the SpaP (antigen I/II) polypeptide, M and M-like cell-surface proteins from Streptococcus pyogenes, and cell wall Protein A from Staphylococcus aureus, revealed that, in addition to these proteins having N-terminal leader sequences to direct their secretion, they carried a common motif at the C-terminal end. The presence of this motif, initially identified as the amino acid sequence Leu-Pro-Xaa-Thr-Gly-Xaa (where Xaa is any amino acid), is now recognized as a specific sequence that directs anchorage of proteins to the Gram-positive bacterial cell wall. Thus it became possible to assign cell wall location to a polypeptide based simply on the presence of this motif. Various other signature motifs were also identified within other secreted proteins that bound fibronectin or polysaccharides. These assignments, confirmed by experimental data, enhanced general understanding of cell surface protein structure and function. There are many examples of protein function being inaccurately assigned on the basis of sequence analysis without supporting experimental evidence. However, what's in a sequence can lead as well as mislead. In a development that happened over several years, the function of a streptococcal surface protein family came to be assigned correctly on the basis of initial sequence analysis. By 1992, it was known that streptococci expressed a class of cell surface proteins that were not cell-wall-anchored, but were tethered to the cytoplasmic membrane (Jenkinson, 1992). These so-called "lipoproteins" could be identified through possession of an N-terminal amino acid motif Leu-Xaa-Ala/Gly-Cys (where Xaa was usually a hydrophobic residue). We were interested in determining the function of a family of 35-kDa polypeptides (designated LraI), members of which had been identified in 4 species of streptococci. These proteins had been tentatively assigned adhesive functions by three independent laboratories (Jenkinson, 1994). The deduced amino acid sequences of the LraI polypeptides contained the lipoprotein modification motif, but it was not easy to envisage how these polypeptides, which were presumably bound close to the cytoplasmic membrane, could function as adhesins. By sequencing of the DNA surrounding the ScaA lipoprotein gene in S. gordonii, Paul Kolenbrander's laboratory at NIDR showed that scaA constituted part of an ATP-binding cassette transporter operon. Paul decided to come to Otago, New Zealand, in 1996 so that we could work together on Sca. We managed to obtain experimental data that demonstrated conclusively that ScaA, initially believed to be a co-aggregation adhesin, was in fact a component of a transport system associated with manganese uptake (Kolenbrander et al., 1998). Subsequently, it has been verified that some of the other streptococcal LraI polypeptides are also involved in metal ion uptake, an essential function for growth and survival of bacteria in the host. To my mind, there was an epic discovery made in 1995 for streptococcal research. The studies by MacLeod in the 1950s, and by Pakula and Walczak in the early 1960s, had demonstrated that exponential growth phase cultures of some strains of pneumococci and streptococci could attain a critical state of "competence" during which DNA could be taken up. Work in the late 1960s and early 1970s had established that competence for transformation could be induced by a proteinaceous factor (known as competence factor, CF) present in the extracellular fluid of competent cultures. This factor had never been purified or characterized. In 1995, it seemed that almost overnight the mystique of competence development was resolved. Morrison (Chicago) and Havarstein (Norway) had embarked upon a collaborative project that led to the purification of a 17-amino-acid-residue competence factor from pneumococcus and identification of the gene encoding the precursor (Havarstein et al., 1995). By then sequencing the surrounding DNA, they identified the genes encoding components of a signal transduction system that sensed competence factor and controlled competence development, in pneumococcus and in S. gordonii. The competence regulons appear to be intricately linked with the lifestyle of the bacteria, because the CF sensor-regulator system is essential for bacterial survival in the animal host.
There was another milestone event in 1995 that was perhaps more of an accomplishment than a discovery. This was the publication of the first complete nucleotide sequence of a microbial genome, Haemophilus influenzae. There are now about 50 microbial genomes fully sequenced, with many more in progress. Although a bacterial genome can be sequenced in just a few hours, there are still information bottlenecks associated with sequence annotation. Thus, the genome sequences for oral bacteria P. gingivalis, A. actinomycetemcomitans, Treponema denticola, and S. mutans are mostly completed but not yet fully annotated. Other oral bacterial genomes that will be sequenced in the next few years include those from S. sanguis, S. gordonii, A. naeslundii, Fusobacterium nucleatum, and Bacteroides forsythus. Comparative genomic studies of species and strains are helping to build pictures of how genomes evolve. However, for functional studies, these sequences will provide us with the opportunity to define the complements of genes expressed in association with phenotypes, responses, or lifestyles. For example, when the relevant gene microarrays become available, we will be able to investigate the gene expression profiles of S. mutans cells growing at pH 5, and of P. gingivalis growing within a periodontal pocket or epithelial cell. Advanced studies will allow for identification of the "interactome"–the sets of proteins with which each gene product physically interacts. The "adhesiome", "biofilome", and "coaggreome" may provide gene expression catalogs associated with oral bacterial interactions with host cells, surfaces, and other bacteria. Although the polypeptides encoded by between 30% and 50% of open reading frames (ORFs) within sequenced genomes currently have no functional identities, some of these will undoubtedly be resolved through microarray analyses, since they will be co-ordinately expressed with known genes. However, to determine functions for many of these ORFs, investigators will still find it necessary to express and purify the gene products, inactivate the genes, and phenotype the mutants. This seems remarkably like 1980, where I came in, so I feel at least partially equipped to deal with the immediate future challenges of the new era of oral microbial genomics. Received for publication November 27, 2001. Accepted for publication December 18, 2001.
Journal of Dental Research, Vol. 81, No. 2,
84-88 (2002)
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
INTRODUCTION
JOURNEY INTO ORAL MICROBIOLOGY