Abstract
For more than 100 years, investigators have tried to identify the bacteria responsible for dental caries formation and to determine whether their role is one of specificity. Frequent association of Lactobacillus acidophilus and Streptococcus mutans with caries activity gave credence to their being specific cariogens. However, dental caries occurrence in their absence, and the presence of other bacteria able to produce substantial amounts of acid from fermentable carbohydrate, provided arguments for non-specificity. In the 1940s, Stephan found that the mixed bacteria in dental plaque produced a rapid drop in pH following a sugar rinse and a slow pH return toward baseline. This response became a cornerstone of plaque and mixed-bacterial involvement in dental caries causation when Stephan showed that the pH decrease was inversely and clearly related to caries activity. Detailed examination of the pH (acid-base) metabolisms of oral pure cultures, dental plaque, and salivary sediment identified the main bacteria and metabolic processes responsible for the pH metabolism of dental plaque. It was discovered that this metabolism in different individuals, in plaque in different dentition locations within individuals, and in individuals of different levels of caries activity could be described in terms of a relatively small number of acid-base metabolic processes. This led to an overall bacterial metabolic vector concept for dental plaque, and helped unravel the bacterial involvement in the degradation of the carbohydrate and nitrogenous substrates that produce the acids and alkali that affect the pH and favor and inhibit dental caries production, respectively. A central role of oral arginolytic and non-arginolytic acidogens in the production of the Stephan pH curve was discovered. The non-arginolytics could produce only the pH fall part of this curve, whereas the arginolytics could produce both the fall and the rise. The net result of the latter was a less acidic Stephan pH curve. Both kinds of bacteria are numerous in dental plaque. By varying their ratios, we were easily able to produce Stephan pH curves indicative of different levels of caries activity. This and substantial related metabolic and microbial data indicated that it is the proportions and numbers of acid-base-producing bacteria that are at the core of dental caries activity. The elimination of S. mutans, as with a vaccine, was considered to have little chance of success in preventing dental caries in humans, since, in most cases, this would simply make more room for one or more of the many acidogens remaining. An understanding of mixed-bacterial metabolism, knowledge of how to manipulate and work with mixed bacteria, and the use of a bacterial metabolic vector approach as described in this article have led to (1) a more ecological focus for dealing with dental caries, and (2) new means of developing and evaluating anti-caries agents directed toward microbial mixtures that counter excess acid accumulation and tooth demineralization.
Research on dental caries has a long history, and the biology of its
development has proved to be far more complex than anyone might have thought.
Specificity and non-specificity of the causative bacteria have been at the
center of the controversy of how this disease occurs and how it should be
treated. From this article, it appears that the composition, metabolism, and
caries-causing capability of applicable bacterial mixtures are integrated, and
that they are affected collectively by oral ecological conditions and changes.
This is an area of Oral Biology that needs further exploration and could
provide significant dividends in the development of our understanding of human
ecological infections in general.
–Olav Alvares, Editor
(I) Introduction
These studies enabled Miller to formulate his acid decalcification theory for dental caries which, in simple terms, involved two steps: First, the mixed bacteria resident in the mouth produce acid from fermentable carbohydrate; and second, the acid then dissolves tooth mineral to initiate and sustain the cavity-producing process. Stimulation of acid production has been extensively studied because of its central role in the caries process, and was confirmed many times with bacterial isolates from the oral microbiota (Stephan and Hemmens, 1947; Kleinberg et al., 1982; Denepitiya and Kleinberg, 1984; Wijeyeweera and Kleinberg, 1989a; van Houte, 1994; van Houte et al., 1994, 1996), with dental plaque and salivary sediment in vitro (Kleinberg et al., 1982; Singer et al., 1982; Salako and Kleinberg, 1992), and with plaque in vivo (Stephan, 1940, 1944; Strålfors, 1950; Kleinberg, 1961; Imfeld, 1983).
Since the time of Miller, there has been continual controversy regarding the role of bacteria in dental caries causation, particularly in humans. At issue have been (i) whether a specific bacterial species or a non-specific mixture of bacteria is the agent responsible, and (ii) whether it is an infectious bacterial disease in the classic sense or an ecological overgrowth.
In recent years, Loesche (1986) brought focus to this controversy by proposing a specific-/non-specific-plaque hypothesis for dental caries and concluded that specificity and classic infection were on the correct side of the debate. Accordingly, he argued that the way to deal with the microbial aspect of the dental caries problem is targeted elimination from the oral microbiota of what is generally known as Streptococcus mutans, or the mutans streptococci (for example, with antibiotics or vaccines; Loesche, 1986; Taubman and Smith, 1993). In framing the argument, Loesche considers that those favoring the non-specific side of his hypothesis believe that the oral bacteria as a whole produce the acid from fermentable carbohydrate that is central to dental caries formation. So the role of bacteria in caries causation is one of non-specificity. In this case, removal of as much as possible of the plaque that accumulates on the teeth is a logical means of dealing with the caries problem.
To a certain degree, the author of the present article falls into the non-specific group, since he has concluded that Streptococcus mutans cannot be, except in a few specific cases, the bacterial cause of dental caries. This is simply because many non-mutans micro-organisms are sufficiently acidogenic/aciduric and numerous to produce the amounts of acid necessary (van Houte, 1994, and others). However, and as covered below, he also considers the bacteria involved in dental caries causation to be due to an increase in a mix or spectrum of resident acidogenic/aciduric oral micro-organisms and/or a decrease in a mix or spectrum of resident bacteria best able to produce counteracting alkali. Like others, he considers such microflora changes reactive, since they occur in response to changes in the oral ecology (Marsh, 1989).
The present article identifies problems with S. mutans being the causative element of the dental caries process and has as its main thrust the thesis that dental caries results largely when the bacterial compositions of the microflora resident in predominantly retentive dentition sites are such that they generate excessive amounts of acid and/or deficient amounts of base. As I hope will become evident below, alkali deficiency is an important element of the microbial dysfunction involved, and bacterial alkali production is closely linked to the protection against caries afforded by saliva.
In accordance with the Loesche hypothesis, which considers certain of the mutans streptococci to be the primary cause of dental caries, treatments and cures are directed toward elimination of these micro-organisms. On the other hand, where mixtures of diverse acidogens are involved, and they arise in response to ecological change, treatments and cures need to be directed toward correction of, or compensation for, the oral ecological dysfunction involved. The bacteria, while essential, are necessarily secondary. In this regard, encouraging formation of alkali by the oral bacteria is likely to be an important aspect.
(II) Lactobacilli and Streptococci as Specific Dental Caries Infectious Agents
The lactobacilli and streptococci are major genera of the category of bacteria generally referred to as the lactic acid bacteria (Thompson, 1987), so it is no surprise that microbial species within these genera have been proposed as specific agents of the acid production that is primary to the dental caries process (Hamada and Slade, 1980; Loesche, 1986; van Houte et al., 1994; Liljemark and Bloomquist, 1996). In a search for a caries-causative micro-organism from among the mixed oral microbiota, Bunting et al. (1929) and Jay (1947) identified lactobacilli in general and Lactobacillus acidophilus (B. acidophilus) in particular as possible candidates. Many studies carried out by them and thereafter by others have shown frequent association between the presence of lactobacilli and the prevalence of dental caries, suggesting such a possibility. For example, elevation of the level of fermentable carbohydrate in the diet led to elevated lactobacillus counts, whereas lowering of such carbohydrate resulted in lactobacillus reduction (Becks et al., 1944; Becks, 1950). Dentition sites favoring retention of fermentable dietary carbohydrate also favored elevated numbers of lactobacilli and the development of dental caries lesions (Stecksén-Blicks, 1985; Crossner et al., 1989). These sites included the pits, fissures, and approximal areas of the teeth where caries lesions are most frequently found (Klein and Palmer, 1941; Barr et al., 1957). Also observed was that the placement of dental appliances such as orthodontic bands on dentition sites changes the morphological conditions, which then lead to enhanced carbohydrate retention, more lactobacilli and other acidogens, a more acidogenic dental plaque, and, in turn, to caries elevation (Sakamaki and Bahn, 1968; Balenseifen and Madonia, 1970; Chatterjee and Kleinberg, 1979; Scheie et al., 1984; Boyar et al., 1989).
Despite these relationships, other observations indicated that lactobacilli are not essential for caries development. This is because caries lesions can develop in the absence of lactobacilli, and other acidogens that are resident members of the oral microbiota can provide the necessary acid. The relationship between lactobacilli and dental caries, at least in humans, was not proven to be cause-and-effect. A better argument could be made for its being associative (Sims, 1985).
The intensity of this difference in belief subsided until the early 1960s. It then flared up again with S. mutans (first isolated by Clarke in 1924 from caries lesions) being identified as a possible caries-causing candidate (Fitzgerald and Keyes, 1960; Keyes, 1960; Keyes and Jordan, 1963; Carlsson, 1968; Guggenheim, 1968; Duchin and van Houte, 1978). To the large majority of researchers in this field, it became the caries cause célèbre, and a massive amount of research ensued. Subsequent studies led to the discovery of at least 8 serotypes of S. mutans, which were then grouped mainly into 4 species relevant to humans and small laboratory animals (Loesche, 1986; Michalek and Childers, 1990). Serotypes c, e, and f remained S. mutans; d, g, and h became S sobrinus; serotype a became S. cricetus, and serotype b became S. rattus. Collectively, they are now referred to as the mutans streptococci. To other investigators, including the author, S. mutans has simply been another acidogenic micro-organism of many within the human mouth that are able to produce relatively large amounts of acid. As indicated above, acid is the central virulence agent in dental caries development, and, as seen earlier for lactobacilli, similar associations with availability of fermentable carbohydrate in the diet and with dental caries were also observed (see Kleinberg, 1977a,b, for references and discussion; Hamada and Slade, 1980; Loesche, 1986). In a sense and to a large degree, history repeated itself with the “discovery” of this new microbial acidogen.
(III) Basis and Dynamics of Acid-Base pH Change in Dental Plaque in vivo
In the interim period between focus on species of lactobacilli, particularly L. acidophilus, and species of streptococci, particularly S. mutans, significant advances were made in identifying their roles in the dental caries process of dental plaque as a whole and a tendency to move in the direction of less specificity. This resulted from two landmark investigations, carried out by Stephan (1940, 1944), demonstrating that dental plaque has the ability to produce rapid and substantial decreases in pH in vivo. The initial study showed for the first time that dental plaque that was allowed to accumulate on readily accessible incisor teeth could produce a rapid and substantial decrease in the pH immediately following exposure to a sugar challenge in the form of a rinse with a glucose or sucrose solution. After reaching a minimum, the pH showed a subsequent slow rise to baseline, which usually took about one hour. The second study was of even greater importance, since it showed that the extent of decrease and location on the pH scale of this sugar-challenge curve (subsequently called the Stephan curve) was inversely related to the caries activities of the subjects tested (Fig. 1). He measured caries activity by scoring individuals for caries lesions over a two-year period and establishing subject groups with caries activities ranging from caries-free to highly caries-active (Fig. 1).
In a subsequent study in vivo, Kleinberg (1961) manipulated sugar substrate availability in a wide variety of ways and showed a clear cause-and-effect relationship between substrate availability and the magnitude and duration of the subsequent plaque pH fall (Fig. 2). Glucose was the substrate provided, and availability was defined as the product of substrate concentration and the time it was available. The pH fall, the reaching of a pH minimum and the slow pH rise thereafter, and the characteristics of the Stephan curve were explained in terms of the kinetics of acid production and acid removal under open system conditions and conditions of limited availability of sugar substrate. Delay in the return of the pH to baseline was observed and explained by the prolongation of acid production when the availability of glucose substrate was prolonged. The extent of the plaque pH decrease (i.e., how low and for how long) was attributed to the concentration of bacteria present and the carbohydrate substrate available (Kleinberg, 1961, 1970a).
That the basis of the Stephan curve was essentially an enzyme-substrate type of relationship expressed in an open system was supported in a follow-up study with urea (Kleinberg, 1967b). This showed that making this nitrogenous substrate available to plaque in situ, as in the glucose study, gave almost exactly the same results, except that base formation and rise in the pH occurred rather than acid formation and a fall in the pH (Fig. 2). The results with glucose and urea were basically symmetrical and indicated that both acid and base formation can contribute to the plaque pH and serve as counteracting metabolic forces in the caries process.
The two Kleinberg plaque in vivo studies (1961, 1967b) led to the classification of plaque pH responses into two types and the recognition that they were determined by the substrate provided and its availability. They also led to the vector format shown in Fig. 3, where the effects on the plaque pH of low and high substrate availability can be seen. With glucose as substrate, the first condition produced a pH curve in which there was a rapid pH fall and a subsequent slow pH rise which occurred once the glucose was consumed by the plaque bacteria (i.e., the Stephan pH curve response). The second (high substrate) condition also produced a curve with a rapid pH fall, but this was followed by a “bottoming out” of the pH, where it remained as long as there was fermentable carbohydrate substrate still available. In the case of urea, it was pH rise and fall, and pH rise and plateauing, respectively.
What was striking in these Kleinberg pH studies was that dental plaque, a complex mixture of several hundred species of resident bacteria (Moore et al., 1982; Liljemark and Bloomquist, 1996), functioned as a single entity and did so in a relatively simple and predictable manner. This led to the recognition that pH could be used as a master variable of this mixed-bacterial system (Kleinberg, 1970a), and that pH response to substrate challenges with primary substrates such as glucose and urea was an easy way to measure the dynamics of overall pH (acid-base) metabolism in real time.
(IV) Dental Plaque as a Mixed-bacterial Entity of Coordinated pH Metabolisms
In an extensive survey of the pH of dental plaque in different dentition locations throughout the mouth and its relation to meals, rates of resting salivary flow, and salivary pH, an observation of broad significance was made (Kleinberg and Jenkins, 1964). Plaque in different dentition sites within a given person's mouth, in the same dentition sites in different individuals' mouths, and in sites showing different caries prevalence rates all showed the same type of pH response after eating, except that the responses varied in magnitude of pH change and location on the pH scale (Fig. 4a). In this study on 85 subjects (and about 12,000 pH measurements), the pH was highest in the morning before the subjects ate breakfast, lower after fermentable carbohydrate became available to the plaque bacteria throughout the mouth, and rose slowly thereafter toward baseline over a mean period of 3½ hours. The site variation in this pH synchrony was attributed by Kleinberg and Jenkins (1964) to two primary factors. One was variation in the access of the dentition sites to saliva, which followed a pattern largely determined by variation in amounts of saliva from the three pairs of major salivary glands and the intra-oral locations of their duct orifices (Schneyer and Levin, 1955a,b; Sreebny, 1987, 2000). The other was attributed to variations in tooth and dentition morphology (Kleinberg, 1978). These variations affect the amounts and types of bacteria that are able to accumulate in the different tooth and dentition sites throughout the mouth, and substrate availability. Both are primary to acid formation: One supplies bacterial enzymes, and the other the substrates.
A significant observation made in the comprehensive Kleinberg and Jenkins (1964) study with broad consequences was that plaque in dentition sites favored by saliva, and in individuals with a higher basal (i.e., resting) rate of salivary flow and pH, favored higher levels of plaque pH (Fig. 4b). Note that the resting saliva pH fell between the fasting (i.e., highest) pH and the minimum (i.e., lowest) pH of the plaque pH ranges shown. Its significance is that a sustained fasting pH so much higher than the salivary pH, as seen here, can happen only if substantial base formation is continually taking place within the dental plaque, particularly at this time of the day (Kleinberg and Jenkins, 1964; Kleinberg, 1967b).
The synchrony seen in the Kleinberg and Jenkins investigation is illustrated in Fig. 5, where the pH of dental plaque of three representative dentition sites was compared after two successive sugar challenges of different magnitudes. The influence of saliva is relevant when maxillary and mandibular interproximal plaque samples are compared (also cf. Fig. 1). The effect of plaque thickness, to a large degree, is probably involved when maxillary labial and maxillary interproximal plaque samples are compared. Important to note is that the pH changes in these three dentition locations were in synchrony, but the locations on the pH scale and/or their magnitudes of pH response were different, presumably because of differences in saliva availability and amounts of plaque that accumulate because of dentition location and morphology.
(V) Simulation of the Stephan pH Curve in vitro with Single and Mixed-bacterial Populations
As a first step toward identification of the bacteria mainly responsible for the Stephan curve and its simulation, Stephan and Hemmens (1947) prepared bacterial isolates from dental plaque and examined each alone and in mixtures to see whether and which could give in vitro the rapid pH fall and subsequent slower pH rise characteristic of the Stephan curve in vivo. Alone, pure cultures of many of the oral bacteria tested could produce an acidic pH response, but few could give the curve seen in vivo for plaque as a whole. In the experiments in which they explored mixtures of the isolates, they opened the door to the possibility that bacterial mixtures as well as single micro-organisms might be Stephan pH curve simulators. It soon became clear that mixing pure cultures to identify the bacteria responsible for production of the Stephan curve and its relation to dental caries activity was not an easy task. It meant sorting through so many combinations and permutations of plaque bacterial isolates that the task of finding bacterial mixtures or combinations that were applicable was extremely remote. Combining more than a few bacteria can be an experimental nightmare (see, for example, Fig. 9 of Wijeyeweera and Kleinberg, 1989b). These difficulties may have been the reason for this important work of Stephan and Hemmens (1947) being hardly noticed and coming to an end.
Despite their inability to identify the bacteria primarily responsible for the Stephan curve, Stephan and Hemmens (1947) nonetheless made several important observations, one of which was that a high concentration of oral acidogenic micro-organisms was essential for the production of the initial rapid and oftentimes extensive pH fall seen in plaque in vivo in response to a sugar substrate challenge. Using pure culture biofilms, Strålfors (1950) confirmed the importance of high bacterial cell concentrations in the production of rapid acidification of the plaque pH and in the formation of the Stephan curve. The biofilms he constructed consisted of pure cultures of oral bacteria encapsulated in agar gels formed around a glass pH electrode in a configuration that simulated dental plaque in situ and enabled the pH of both its inner and outer surfaces to be measured at the same time. Thereafter, Kleinberg (1967a) showed that the Stephan pH curve could be simulated with salivary sediment suspensions. These various studies were collectively extremely important, because they demonstrated that this curve could be modeled with bacteria either in suspensions or in biofilms, with pure or mixed bacterial cultures, and either in vivo or in vitro.
Subsequent studies in our laboratory proved salivary sediment to be metabolically and microbially remarkably similar to pooled dental plaque (Kleinberg et al., 1982; Singer et al., 1982; Salako and Kleinberg, 1992; Ryan and Kleinberg, 1995a). Salivary sediment is easily obtained from stimulated whole saliva by centrifugation. It consists of a sampling of the bacteria that are ready to be dislodged or shed from the surfaces of the oral hard and soft tissues (Kleinberg and Westbay, 1992; Liljemark and Bloomquist, 1996). The bacteria on these surfaces constitute a continuous culture of mixed bacteria, and many are shed as part of mucosal epithelial desquamation. A major advantage of using salivary sediment as a simulant of dental plaque acid-base metabolism is that it is available in abundance and provides a sampling of oral mixed bacteria “on tap” that have grown in and are compatible with oral environmental conditions. There is no need to grow oral isolates outside the mouth and recombine them for use in many oral mixed-flora experiments. In the use of salivary sediment for study, it was found important that the cell concentration be appropriately adjusted and fixed (Singer et al., 1982), and that donor subjects had fasted overnight before saliva collection was made in the morning (Sandham and Kleinberg, 1969). Overnight fasting ensured that little or no bacterial carbohydrate could be stored to affect the initial pH and the extent and nature of the pH fall in these experiments (Denepitiya and Kleinberg, 1984). For many experiments, a pre-incubation to deplete such substrates was also done.
(VI) Determination of the Main Metabolic Pathways and Acids Involved in the Lowering of the pH when Fermentable Carbohydrate is Catabolized by the Oral Mixed Bacteria
A series of investigations was carried out to characterize the metabolism of the salivary sediment system under glucose substrate conditions that produced the main types of pH curves, like those seen in plaque in situ (Sandham and Kleinberg, 1969, 1970a,b). Along with standardization of the sediment concentration at the 16.7% (v/v) level, the glucose concentration was varied between 0 and 30% (w/v), as was done in vivo (Kleinberg, 1961), and pH, uptake of glucose, and formation of lactic and other short-chain carboxylic acids (SCCA) were determined along with carbon dioxide over a four-hour period of incubation at 37°C. At the same time, experiments were done where the carbon dioxide produced from glucose labeled on different carbon atoms with 14C was examined by a micro-radiorespirometric method for determination, by 14C tracing, the metabolic pathways whereby glucose is degraded (Sandham and Kleinberg, 1970b).
These experiments showed progressively greater acidic pH responses as the glucose concentration was progressively increased. An active Embden-Meyerhof Scheme (EMS) and an incomplete tricarboxyllic acid (TCA) cycle were demonstrated. Instead of an intact cycle, pyruvate arising out of the EMS pathway flows into different types of short-chain carboxylic acids, with the types and their proportions determined by the glucose level. At low glucose levels, when the pH rapidly fell and then slowly rose, L(+) lactic acid correspondingly rapidly rose and then slowly fell (Sandham and Kleinberg, 1970a; cf. dental plaque in vivo in Geddes, 1972, 1975). Basically, it behaved like a metabolic intermediary, since it is subsequently converted mainly to carbon dioxide and acetic acid. In contrast, at high glucose levels, when the pH fell and reached and remained at a more acidic level, lactic acid rose asymptotically and behaved like an end-product along with acetic and propionic acids, which progressively rose (Sandham and Kleinberg, 1970a). An important factor that comes into play at high sugar concentrations and affects the SCCA proportions is that a low pH is reached which can inhibit or slow lactic acid formation (Iwami and Yamada, 1980), and its subsequent conversion to acetic and propionic acids is slowed or stopped (Sandham and Kleinberg, 1970a).
An important observation was that salivary sediment, even though comprised of many different types of bacterial cells, evidently behaved in synchrony, and the study of its overall metabolism was no more difficult than studying pure cultures or simple pure culture mixtures (Traudt and Kleinberg, 1999). Simplified for subsequent experiments was the observation, once more, that two kinds of pH curves are observed as in plaque in vivo, one at low sugar availability (initial fast pH fall followed by a slow pH rise) and the other at high sugar availability (initial rapid pH fall followed by a delayed or no pH rise). Explanations for pH curve shapes and their relation to glucose utilization and amounts and types of short-chain carboxylic acids generated were made evident by these experiments (Sandham and Kleinberg, 1970a). Also evident from these studies was that the metabolic configuration of the sediment microflora entity was an aerobic glycolysis type of metabolism (Korayem et al., 1990), which fits in with the low oxygen environment found in the mouth (Globerman and Kleinberg, 1979).
(VII) Effects of Saliva on the Stephan pH Curve
Saliva is the main biological fluid bathing the oral microbiota in the oral cavity and is the medium of the salivary sediment system. My colleagues and I therefore performed experiments to examine its effects on the Stephan pH curve in the sediment system and on the fermentation parameters responsible for its shape (Kleinberg et al., 1973). The effects of saliva were profound (Fig. 6), and its manifestations at low and high substrate levels were different. At low substrate concentrations, there was a less-than-optimal pH fall with saliva present; the pH reached a minimum, and then subsequently slowly rose toward baseline. When saliva was absent, the initial pH fall was rapid as before but instead fell to a lower level, reached a pH where it leveled off, and remained there for much of the experimental period. In other words, the pH fall without saliva was enhanced, and the rise part of the Stephan pH curve was lost. Part of the pH rise loss may be due to the absence of salivary bicarbonate (Wijeyeweera and Kleinberg, 1989b).
At high substrate concentrations with substrate in excess, the pH fell as before, and no pH rise was seen with or without saliva (Fig. 6). However, examination for glucose uptake and other metabolic parameters showed that glucose utilization and acid formation were much greater with saliva present (Fig. 6). Evidently, base formation and saliva buffering when saliva is present prevent the pH from falling to a lower pH than would otherwise have occurred. The net result is that the glucose substrate is cleared more rapidly because of the saliva. In vivo, clearance is enhanced in those oral sites where saliva is continually being replenished (Dawes, 1989). It appears that excessive fermentation and deficient saliva are two overall factors that can work together to make return of the pH from acidity more difficult than with either alone.
(A) Identification of the pH-rise factor in saliva
To identify the constituents responsible for the pH-raising effect of saliva seen in vitro, we next carried out saliva fractionation experiments. This proved to be mostly arginine available from small arginine peptides (Kleinberg et al., 1979). This common amino acid, especially if supported by saliva buffering, proved to be highly effective (Wijeyeweera and Kleinberg, 1989a). The buffering capacity of saliva is a property that has often shown an inverse correlation with caries prevalence (Afonsky, 1961) and is largely due to salivary bicarbonate (Lilienthal, 1955).
The metabolic pathways involved in arginine degradation were also explored. Its conversion to ornithine was shown to occur via the arginine deiminase pathway (Kanapka and Kleinberg, 1983), and although ornithine can lead to some putrescine formation, subsequent examination indicated that its main degradation occurred via glutamate and α-ketoglutarate to yield, ultimately, ammonia, carbon dioxide, and acetate (Traudt and Kleinberg, 1999).
(B) Contribution of urea to base formation
Urea is also involved in base formation by saliva. When its degradation and effects on the pH were explored in the salivary sediment system, it had minimal effects if no exogenous urea was added. This is because the amounts present in saliva supernatant, when used in this in vitro system, are limited. In contrast, in vivo, where saliva is continuous and the supply of urea is therefore substantial, the effect of urea on the pH is much greater (Kleinberg and Jenkins, 1964), and its effect on caries activity can be considerable (Peterson et al., 1985; Meyerowitz, 1993). For this reason, our studies on urea have been both in vitro and in vivo (Biswas and Kleinberg, 1971; Singer et al., 1982; Singer and Kleinberg, 1983a,b). The experiments on urea catabolism showed mostly production of ammonia and carbon dioxide, but some amino group incorporation into amino acids was also demonstrated (Singer and Kleinberg, 1983a,b). The incorporation was mostly into alanine, especially if there was some glucose present.
(VIII) Microbial Elements Responsible for the Stephan Curve and its Relation to Dental Caries
Once arginine was identified, this amino acid was all that was needed, together with glucose, to produce the Stephan pH curve in vitro. When provided as its bicarbonate salt to ensure buffering like that normally present in saliva (cf. Stephan and Hemmens, 1947), the stage was set to tackle the question of which bacteria of the resident oral microflora entity could produce Stephan pH curves in vitro with only glucose and arginine as the substrates provided (Kleinberg et al., 1982; Wijeyeweera and Kleinberg, 1989b). In other words, the Stephan pH curves seen with and without saliva could now be simulated without the complications of using saliva supernatant, with its multiplicity of substrates and buffers that might affect the pH.
The results were dramatic (Table) and are illustrated for some bacteria in Fig. 7, along with results where arginolytic and non-arginolytic bacteria were combined. Many of the major acidogenic bacteria in the survey produced a pH fall but little or no pH rise (the non-arginolytics), whereas similar acidogenic bacteria also produced a pH fall but one that was less, because these bacteria could produce base from arginine (arginolytics). Mixing the two types of bacteria generally resulted in curves intermediate in value.
Relevant to the dental caries process is that the bacterial members that carry only the acid-forming function are those that most tend to be cariogenic. When fermentable carbohydrate becomes available in the mouth during eating, one can expect these bacteria to produce a substantial amount of the acid, cause a rapid fall in the pH, and keep the pH acidic for as long as there is fermentable carbohydrate left to sustain acid production. In contrast, the bacteria in plaque that also possess the ability to raise the pH (by degrading arginine and producing base; Kanapka and Kleinberg, 1983) would be expected to produce less net acid and as a consequence be less or non-cariogenic. Hence, it is interesting to note that all of the mutans streptococci, except for S. rattus, are members of the first group, whereas S. rattus belongs to the second group (Table). The same two kinds of acid-producing bacteria are seen with the oral lactobacilli. Some are arginolytic, while others, such as L. acidophilus and L. casei, are not (Table). In other words, the mutans streptococci and lactobacilli that have been related to increased dental caries activity (Hamada and Slade, 1980; Loesche, 1986; and others) are species that are non-arginolytic, whereas S. rattus, which has been associated both in animals and in humans with the caries-free condition (Kilian et al., 1979), is arginolytic.
Evident from these experiments was that a higher ratio of arginolytic to non-arginolytic bacteria in a mixture of the two is conducive to a less cariogenic type of Stephan pH curve, whereas the reverse ratio is more favorable to a more caries-active type of curve. At long last, we were able to achieve what Stephan and Hemmens (1947) had hoped to do more than 50 years ago. Also evident is that acid and base formation both occur in dental plaque and counteract each other in the determination of plaque pH. As a consequence, a deficiency in base formation (largely associated with saliva) can be as important in dental caries development as can excessive formation of acid from fermentable carbohydrate (largely associated with the diet). Identification of arginine as a means of combating dental caries by enhancing base formation has added another and important basis for saliva being protective against this disease. Urea was identified earlier as a possibility for enhancing the formation of base, but it can have some problems if inappropriately used as an anti-caries agent (Jenkins and Wright, 1950, 1951).
Further experiments enabled flora shifts associated with greater acidogenicity and caries development to be simulated and thus effects on the pH curve to be studied. For example, change from a non-cariogenic to a cariogenic type of microflora is favored when the oral environment is enriched with fermentable carbohydrate from the diet (Becks et al., 1944), or when a condition of saliva deficiency develops, as happens with onset of a xerostomic condition (Sreebny, 1987, 2000). Simulation in vitro is accomplished by the addition, to plaque or sediment, of the non-arginolytic, acidogenic bacteria that might rise when such a shift occurs (Fig. 8). The converse would simulate a change to a non-cariogenic microflora by the addition of an arginolytic micro-organism (Fig. 8). This is a simple means of flora manipulation that opens the way for quantitative study of the effects on the acidogenicity such changes could or would produce. It identifies as well the influence, on the metabolism of the microflora entity, of changes in the proportions of its microbial constituents, whether caused experimentally or ecologically.
(IX) Bacteria that are Significant Contributors to the Biochemical Processes that Produce the Plaque Acid-Base Metabolic Vectors
Extensive studies have been carried out in our laboratory to identify which biochemical processes and which of the numerically more prominent oral bacteria are responsible for the metabolic vectors that form the foundation of plaque pH metabolism (Fig. 9a). From these extensive microbial-metabolic investigations, it was found that most Gram-positive bacteria and some Gram-negatives contribute to vector 3 (Salako and Kleinberg, 1992; Ryan and Kleinberg, 1995a), which is of prime importance in the causation of dental caries (Stephan, 1944). H. parainfluenzae, A. viscosus, A. naeslundii, and Staph. epidermidis and possibly Strep. salivarius are key contributors to vector 1 (Sissons et al., 1988; Salako and Kleinberg, 1989), which is important for calculus formation and dental caries resistance. The pH-rise bacteria listed in the Table contribute to vector 4, which in this article is considered a key element in counteracting dental caries development. A mix of Gram-positive and Gram-negative bacteria contributes to vector 2 (unpublished data). Some of the short-chain carboxylic acids (SCCA) produced in vector 2 from amino acids are active in eliciting cell and gingival crevicular fluid changes associated with gingivitis (Niederman et al., 1997).
(X) Combination Experiments with Pure Cultures Relevant to Plaque pH Metabolism
Experiments with pure cultures have demonstrated that bacteria carrying one or more of the metabolic vectors shown in Fig. 9a will generally give an averaging effect when similar vectors are combined (see Figs. 7 and 8 ). However, when the bacteria mixed carry different metabolic vectors, then the result is one of addition. This would be the case where there is, for example, a tandem relationship between two micro-organisms. A streptococcus that produces lactic acid from glucose in combination with a veillonella organism that utilizes the lactic acid produced is a good example of oral bacteria in a tandem relationship (Hamilton and Ng, 1983). Another good example is the production of hydrogen peroxide from glucose by any of several streptococci and the utilization of the peroxide they produce by Neisseria sicca, Haemophilus parainfluenzae and segnis, Actinomyces viscosus, and Staphylococcus epidermidis (Ryan and Kleinberg, 1995b).
The benefit of using a metabolic vector approach is that it “bundles” bacteria metabolically and thus makes it much easier to handle bacterial mixtures both conceptually and experimentally. Also, ecological conditions and their relation to disease-causing potential are more easily perceived and symbolized (Fig. 9b; Kleinberg and Westbay, 1992). One can look at mixtures as functional units rather than as a collection of different kinds of bacterial cells.
Manipulations are also relatively easy. For example, one can take pooled plaque or salivary sediment and, despite their cellular complexities, introduce one or more bacteria with a desired metabolic function and in desired amounts to determine if and how much the relevant metabolism of the original microbial composition is altered (Fig. 10; Ryan and Kleinberg, 1995a). In this way, one can assess whether a particular micro-organism can modify the microflora capabilities—for example, whether acid generated from a fermentable carbohydrate substrate can elevate the production of acid to a level critical for the development of caries lesions. By way of illustration, when an acidogen is present at about 1 to 2% or less, its contribution to total acid formation and its impact on the pH would generally be minimal (Sims, 1985; Wijeyeweera and Kleinberg, 1989b). Lower levels than this are the norm for S. mutans in many caries-active situations (Loesche, 1986; van Houte, 1993); this is another reason for considering S. mutans to be a minor organism in human dental caries formation. Many other flora manipulation experiments are possible, but of particular value are modeling experiments for product development, which we have successfully done for the treatment of oral malodor (Kleinberg and Codipilly, 1999).
(XI) Specificity of Streptococcus mutans in Dental Caries Formation
With the re-discovery of S. mutans, a tremendous surge in research occurred during the 1970s and 1980s which appeared to implicate this micro-organism as the caries-causative agent and to establish its specificity. Much of the belief in S. mutans' specificity arose from experimental animal studies which suggested that classic infection was the start of the process. For more than a decade before this, studies had been carried out in an attempt to model caries in laboratory animals, mainly rats and hamsters (Fitzgerald and Keyes, 1960; Keyes, 1960; Keyes and Jordan, 1963; Jordan et al., 1972; Jordan and Sumey, 1973; Tanzer, 1981). At the time, a major frustration for researchers in this area of investigation was that these animals could be caries-prone in one laboratory but not when transferred to another. Loss of such susceptibility continued until Keyes and his colleagues observed that the susceptible animals lost their caries activity when the cariogenic micro-organism(s) was lost from the experimental animals' mouths. However, when it was discovered that S. mutans, either as a pure culture or in plaque from a caries-active individual or even from feces from an infected animal, was introduced into the mouths of such animals, the non-cariogenic deficiency was corrected. Since then, it has become routine practice in animal caries experiments to infect the animal participants with the S. mutans micro-organism.
These findings led to the conclusion that animal caries is basically a classic infectious process, because infection or inoculation by a causative micro-organism starts the process. One needs to keep in mind, however, that sucrose in the diet is usually used to prime the experimental animals (Keyes and Jordan, 1963), which in essence means that the right substrates and the right bacteria are both needed to produce the high levels of acid required for caries generation.
Thereafter, many studies with S. mutans attempted to show that this micro-organism was specific to the caries process and tried to identify how this might be so in humans. Aspects extensively investigated included its ability to: (i) produce acid rapidly from fermentable carbohydrate and lower the pH; (ii) survive and continue to produce acid at acidic pH; (iii) produce high levels of intra- and extracellular polysaccharides, largely as storage components to prolong acid formation and an acidic pH; and (iv) in addition to extracellular polysaccharides, produce polymers to facilitate adherence to tooth surfaces and build-up of large bacterial deposits.
In examining the ability of resident bacterial species to lower the pH and continue to do so at acidic pH, van Houte (1993, 1994) and van Houte et al. (1994, 1996) found that many oral bacteria besides the mutans streptococci and lactobacilli were sufficiently acidogenic to be cariogenic. These other bacteria included non-mutans streptococci such as S. sanguis, mitis, and milleri and bacteria that were neither streptococci nor lactobacilli, such as strains of actinomyces and bifidobacteria. In addition, very low pH levels can be reached with some bacterial mixtures that do not include either S. mutans or lactobacilli. For example, pH levels below 3.5 have been achieved in our laboratory by combining S. oralis with either Neisseria sicca or Veillonella parvula and activating the pH response with relatively high levels of glucose (Traudt and Kleinberg, 1999).
The ability of S. mutans to produce large amounts of extracellular, sticky glucans from sucrose was considered an important part of plaque formation (Gibbons and Nygaard, 1968; Gibbons, 1984, 1989). Adherence of S. mutans was illustrated by its ability to stick to the sides of glass vessels in which these organisms were grown (Gibbons, 1984, 1989) or to wires suspended in a growing S. mutans culture (McCabe et al., 1967). The formation of adherent extracellular dextrans from sucrose was vigorously pursued during this time and was considered to be an essential element in mutans specificity and cariogenicity (see Nyvad, 1993). It was proposed that dextranase, an enzyme able to hydrolyze dextrans, might be used to disperse such gelatinous plaque, but three clinical trials subsequently showed this enzyme to be ineffective (Caldwell et al., 1971; Keyes et al., 1971; Lobene, 1971).
Although bacterial attachment to an acquired pellicle derived from saliva is commonly considered to be the first step in the plaque formation process on the more readily accessible surfaces of freshly cleaned teeth (Egelberg, 1970), other experiments have indicated that bacteria on these and on the poorly accessible dentition surfaces do not necessarily need a pellicle to attach to the tooth surface (Frank and Brendel, 1966). With the development of an intra-oral ultraviolet camera which enables plaque to be visualized without initial staining (Kleinberg et al., 1978), it became patently clear that such a debate might be largely irrelevant to the day-to-day reality where dental plaque spontaneously re-forms when removed by oral hygiene or dislodged by food. This is because photography with this device indicated that the main starting point in most plaque re-formation investigations (that involve no cellular strips) is not formation of a pellicle and then attachment of bacteria at all. Rather, it appears to be the result largely of growth of already-attached bacteria located in the harder-to-clean tooth surface recesses or niches, perhaps with some pellicle material present (Kleinberg, 1987).
This was evident from experiments where even careful removal of plaque by dental prophylaxis (by means of pumice and disclosing solution) still left material visible to the ultraviolet camera but not to the eye (see also Kleinberg, 1977a,b; Gwinnett et al., 1978; Kleinberg, 1978). These remains expanded into larger deposits and progressively spread over the surfaces of the teeth, even though normal toothbrushing was maintained. Scanning electronmicrographs taken in similar experiments indicated that bacterial growth (perhaps with apposition of some bacteria and proteins from the salivary milieu) resulted in the development of small mounds that preceded the spreading out over the tooth surface and in due course becoming a continuous biofilm (Björn and Carlsson, 1964; Kleinberg et al., 1971; Saxton, 1972). Subsequently, Theilade and Theilade (1985) showed, by use of filter-covered films that prevented apposition of bacteria from saliva but not the access of nutrients, that plaque formation is largely due to growth of already-attached bacteria. Growth evidently is the dominant factor, but there are sufficient data to suggest that salivary protein deposits, as well as bacteria from the oral milieu, may make some contribution.
In an attempt to understand how dental plaque is held together, investigators conducted two studies (Silverman and Kleinberg, 1967a,b) in which they harvested plaque from the teeth of human volunteers and, based on earlier work by Dobbs (1932), dispersed it by raising its pH with cold 0.1 M sodium hydroxide. This resulted in the plaque cells and plaque matrix proteins becoming sufficiently negatively charged that they separated, to a large degree, by mutual repulsion (i.e., their zeta-potentials were raised to where they could no longer aggregate). The suspended cellular and solubilized macromolecular components were then separated by centrifugation, and by a series of dialysis and gel filtration steps, the macromolecular components were separated into 11 fractions and analyzed for carbohydrate, calcium, phosphorus, and nitrogen content.
In their experiments, Silverman and Kleinberg (1967b) identified the primary means whereby plaque is held together. At the pH levels seen in the mouth and on oral tooth surfaces, most pH-aggregation titrations indicated that the bacteria have a net negative surface charge. This was later confirmed by Singer (1973) using continuous particle electrophoresis and by Olsson and Glantz (1977) using a similar type of procedure, viz., particle micro-electrophoresis. When the pH, calcium, and the 11 molecular-weight components isolated from dental plaque were examined for their ability to form aggregates with the plaque bacteria, it was evident that calcium ion served as a key cross-linking cation, and that the ability of the bacteria and large-molecular-weight plaque components to aggregate with or without calcium and with or without one another was heavily dependent upon pH. As pointed out by Olsson and Glantz (1977), aggregation and adherence are favored when the zeta-potential of particle surfaces, i.e., bacteria and aggregating macromolecules, approaches zero millivolts. Then, secondary short-range forces and molecules produced or modified by the bacteria can ensure stability to the couplings that have occurred.
(XII) The Tooth-Plaque-Saliva Demineralization-Remineralization Relationship and its Relation to Plaque Acid-Base Metabolism
When Stephan, in his 1944 study, showed different pH levels of response to a sugar challenge by caries-free and caries-active individuals (Fig. 1), analysis of the data immediately suggested the existence of a critical pH or a critical pH range. What this indicated was that tooth demineralization, step two of the Miller acid decalcification process, would not occur unless there was enough acid to decrease the pH to a point where the solubility conditions in the plaque environment became undersaturated with regard to the main mineral of the tooth tissues, viz., the calcium phosphate salt, hydroxyapatite.
Chemical factors preventing teeth from dissolving at a pH above the critical pH level involve two processes. Both are rooted in saliva. One is based on the solubility of tooth enamel in saliva (Fosdick and Starke, 1939). This involves salivary calcium and phosphate ions repressing tooth mineral dissolution by mass action, and, if the pH is sufficiently elevated, replacing lost tooth mineral by remineralization. The other is more complicated and also arises from saliva. It consists of the deposition, into and onto plaque and teeth, of a salivary aggregate consisting of a calcium phosphate carbohydrate protein complex (Fig. 11), which was subsequently named salivary precipitin (Kleinberg et al., 1994).
This complex was fractionated by high-voltage electrophoresis and its constituents analyzed (Chatterjee and Kleinberg, 1979).
Above neutrality, it was high in calcium phosphate (Fig. 11) and is apparently present in an amorphous or poorly crystalline form (Kaufman and Kleinberg, 1973). It is acid-soluble, much more than the calcium phosphate of tooth substance (Kleinberg et al., 1994). This means that when acid is produced by the plaque bacteria from fermentable carbohydrate and the pH drops, the calcium phosphate in the salivary precipitin in plaque, or some modification or alteration thereof, would dissolve before the hydroxyapatite of the tooth (Fig. 12). As a consequence, its role is one of a surrogate source of calcium and phosphate ions, which one might expect would make it possible to maintain calcium phosphate saturation of plaque fluid (Margolis, 1993). This should prevent cavity formation by mass action, as well as provide ions for remineralization to occur. Solubilization of salivary precipitin calcium phosphate can be expected to occur as the pH declines following plaque exposure to fermentable carbohydrate and to cease when the pH rises sometime thereafter, when there is no longer carbohydrate substrate for the plaque bacteria to ferment and generate acid. The pH at which solubilization ceases should be a point of saturation, and the pH between the lowest pH reached during the pH fall and the critical pH should be a zone of undersaturation.
In essence, the acid-base cycling by the plaque bacteria stimulates the demineralization-remineralization cycling among tooth, plaque, and saliva, with plaque as the intermediary (Fig. 13). The role of plaque is complex in that it involves plaque fluid, plaque calcium phosphate, plaque calcium and phosphate ions, and plaque pH in the midst of tooth calcium phosphate, saliva ability to deposit plaque calcium phosphate (Ashley, 1975; Ashley and Wilson, 1977), saliva calcium and phosphate ions, saliva pH, and plaque acid-base metabolism (Fig. 14). Shown in Fig. 14 are the potential plaque and saliva pH levels that could affect tooth and plaque demineralization/remineralization. Generally, caries-free plaque will fluctuate more toward the neutral and alkaline parts of the pH range; highly caries-active plaque will fluctuate more toward the acidic end (see Fig. 1). Saliva pH will range between resting and stimulated pH levels (Dawes and Jenkins, 1964; Jenkins, 1978).
Largely driving these processes are the frequency and duration of the plaque pH cycling periods and periods of saliva replenishment of plaque and tooth mineral. To a large degree, such cycling should eventually determine whether a caries lesion develops (Kirkham et al., 1994).
(XIII) Design of New Anti-caries Compositions Based on These Concepts
Miller's two-stage theory of dental caries initiation and development lends itself to a two-pronged approach to caries prevention and the reversal of any damage caused, at least in the early stages of caries lesion development. To counter the excess acid/deficient base formation component of the caries process (first step), arginine has been selected as the active agent. To counter the excess demineralization/deficient remineralization component (second step), calcium has been selected as the active agent (Kleinberg et al., 1998, 2000). When the two are combined with one or more cariostatic anions, the resulting composition has been named CaviStatTM (Kleinberg, 1999). Bicarbonate/carbonate are preferred anions. Experiments are in progress to evaluate its suitability as an anti-caries agent, alone and in combination with fluoride.
(XIV) Concluding Remarks
S. mutans is still considered by many to be a specific micro-organism that causes dental caries. A large body of research has been developed and assembled that can be taken as support for this thesis (see Loesche, 1986). However, the bulk of the evidence that appears supportive of a relation between S. mutans and dental caries in humans can be considered to be associative and not causative (Sims, 1985). For example, S. mutans often rises in number when there is increased availability of fermentable carbohydrate, especially sugars. Also, it is often (but not invariably) found in those regions of the dentition where caries lesions tend to develop. But unsupportive of its being essential for dental caries causation is that caries lesions can develop in the absence of S. mutans, and S. mutans numbers are often low even when caries is present (Sims, 1985; van Houte, 1993, 1994; van Houte et al., 1994, 1996). As pointed out above, dental caries can develop with other acid-producing bacteria, many of which are present in numbers many-fold higher than those seen for S. mutans. In this regard, the same reasons given before for lactobacilli not being causative also apply to S. mutans. In animals, the situation can be different. There, S. mutans, if of the non-arginolytic type, would be causative if it has to be inoculated for the caries process to begin, and if the oral environment is primed beforehand (usually with elevated levels of sucrose supplied in the diets of the experimental animals). S. rattus, readily found in rats, is arginolytic; it should not be as caries-conducive as the other mutans streptococci, even if sugars were supplied as a priming condition.
With regard to the non-specific part of the Loesche hypothesis, it focuses on the phenomenon that increasing amounts of plaque accumulate on tooth surfaces when oral hygiene is avoided (Ritz, 1967). The increase is spontaneous and includes an increase in the numbers of acidogenic micro-organisms along with other resident oral bacteria. All factors being equal, such an increase in bacterial numbers will give a more acidic Stephan pH response than before the build-up (Stephan and Miller, 1943; Imfeld, 1983). This would make a caries-producing condition possible or an existing caries condition worse.
Removal by physical and/or chemical means of the plaque that has built up reverses the effects of the process and is therefore a logical method of treatment (Stephan and Miller, 1943). Oral hygiene devices, such as toothbrushes with or without toothpastes and the like, would be the order of the day. So might germicides or biocides. Not knowing the causative micro-organism or micro-organisms would not matter (as long as the biocides utilized were not harmful), since removing most of the bacteria in the process would reduce the levels of the caries-causative organism(s) and thus be effective. From the work of Axelsson et al. (1991), frequent removal of plaque at a level necessary to be successful is difficult for a patient to achieve and sustain, but when done in conjunction with professional help, significant caries reduction can occur. The reason for the difficulty is that the pertinent cariogenic oral bacteria are in hard-to-reach dentition sites, viz., pits, fissures, and approximal surfaces of the teeth. Moreover, the oral bacteria grow continuously, and, consequently, their removal needs to be frequent and efficient if the bacterial numbers are to be reduced and kept at levels low enough to ensure that the acid produced is insufficient to produce demineralization and hence caries lesions.
A third approach is indicated by the present article. It is somewhat of a hybrid in that it does not quite fit into either the specific or non-specific aspects of the Loesche hypothesis. First, it involves mixed and diverse bacteria which may or may not include S. mutans. For this reason alone, it has to be considered non-specific. Second, it usually involves ecological overgrowth of certain bacteria and reduction in the numbers of others (e.g., Brown et al., 1975). This unequal effect indicates some specificity and, hence, would exclude it from the non-specific-plaque concept proposed by Loesche. The mixed bacteria that emerge and give this “passive specifity” are those best able to respond effectively to the niche provided by the ecological change.
Ecological changes known to be of prime importance in caries causation include excessive sugar in the diet (Becks, 1950) and significant reduction in the salivary flow (Brown et al., 1975). Both result in a substantial rise in the acidogenic/aciduric bacterial component of the oral flora. Of the two, the salivary change should also result in reduction in the bacteria contributing to alkali formation, and if so, because of the dual effect on the plaque microflora, this might account for the severe caries response often seen in xerostomic patients.
A treatment identified for use in countering the ecological dysfunction involved is the provision of alkali-producing substrates such as arginine and urea (Kleinberg et al., 1979; Rogers et al., 1987; Burne and Marquis, 2000). Arginine is preferred because of the spectrum of bacteria involved (Salako and Kleinberg, 1989) and the practical problems associated with the use of urea. If provided in a dentifrice, such agents would not only favor alkali production to counter oral acidity but should also help, through the physical removal of plaque, to add a further anti-caries benefit by reducing plaque load, which would mean recruitment of the non-specific-plaque hypothesis approach.
Because of the wide scope of the subject dealt with in this article and its focus on the presentation of an alternative to S. mutans and the specific-plaque hypothesis, there are many publications both old and new in an immense body of literature that is relevant but cannot, because of space limitations, be incorporated or referred to here. Nonetheless, what has been framed here should make it possible for the reader to identify where many such articles would apply. The author believes that the concepts and overall thesis presented here should serve as a guide to an understanding of a mixed-bacterial/ecological approach to dental caries causation. It should also be of value in re-interpretation of some of the caries literature, as well as development of an understanding of mixed-flora physiology and how to study it within the oral cavity and in other microbial systems as well.
Relation between the Stephan pH challenge test and dental caries activity in maxillary and mandibular incisor plaque of subjects with different levels of caries activity. The challenge was a 10% (w/v) glucose solution administered as a rinse (25 mL) for two min, and the pH was measured with an antimony-rod touch-type pH electrode (adapted from Stephan, 1944, and reproduced from Kleinberg et al., 1982).
The effect of the availability of glucose and urea substrate on the pH of dental plaque in situ. Availability can be prolonged as shown here by raising the concentration of the challenge substrate or extending the time that substrate is available to the plaque bacteria (from Kleinberg, 1970b).
Vector format of the two types of acid-base pH curves that result from short and long periods of exposure to glucose or urea as seen in Fig. 2. A Type A response results when substrate is limited. A Type B response results when substrate is in excess (adapted from Kleinberg, 1977b).
The effect on plaque pH levels of rinsing for 2 min, first with 100 mL of a 1% and then with 100 mL of a 20% (w/v) glucose solution. Each point is the mean pH of 2 plaque sites in each of five subjects. Each rinse was divided into two 50-mL portions, and rinsing with each was for 1 min (adapted from Kleinberg et al., 1981).
Effects of various levels of salivary supernatant on the pH of salivary
sediment incubation mixtures incubated with
Examples of pH responses of related arginolytic and non-arginolytic bacteria (see Table) mixed in a 1:1 ratio and incubated with glucose at 4.2 mM and arginine at 3.3 mM. Total bacterial cell concentration was 8.3% (v/v). In all cases, the mixing of pure cultures of oral bacteria produced an intermediate result (from Wijeyeweera and Kleinberg, 1989b).
In vitro pH response of pooled dental plaque to which
Catabolism of 2.8 mM glucose when a rapid glucose fermenter, L.
fermentum, is added to salivary sediment in
Relation between the pH of the supernatant of whole saliva (after 10,000 g centrifugation) and aggregation of its constituents. A high-pH plaque favors formation of a higher calcium phosphate and lower carbohydrate protein containing aggregate; a low-pH plaque favors the reverse. Indicated in the Fig. are pH ranges of different kinds of plaque. Possible variation in saliva pH is shown in Fig. 13 (from Kleinberg et al., 1977).
Comparison of the solubilization rates of human tooth enamel slices and salivary precipitin samples prepared from seven different subjects (identified by initials) (from Kleinberg et al., 1994).
Role of saliva in the various calcium phosphate reactions that occur at the
tooth-plaque-saliva interface. Saliva provides calcium and phosphate
Relationship between pH and rate of solubilization of calcium phosphate at the tooth-plaque-saliva interface. Enamel calcium phosphate is mostly hydroxyapatite. Plaque calcium phosphate is x-ray amorphous or poorly crystalline brushite and apatite (Kaufman and Kleinberg, 1973). Since tooth and plaque in vivo are parts of an open system, rates of solubilization are expressed as Ca or P solubilized per unit time. Because plaque calcium phosphate is solubilized more easily than enamel hydroxyapatite, higher plaque fluid levels of calcium and phosphate ions are ensured. As shown, one can expect cariogenic plaque to range over the acidic part of the range shown, calculogenic plaque over the alkaline part, and normal plaque in between (see Fig. 11). Saliva pH covers a narrower range (from Kleinberg et al., 1983).
Footnotes
Acknowledgements
The data in this article come from work carried out in the author's laboratories over a period of more than 40 years, the first three of which were in the laboratory of Dr. Neil Jenkins, under whom the author was fortunate to have done his PhD studies. Dr. Jenkins' collegiality, insights, and depths of knowledge were then, and have since always been, extremely valuable. I would also like to acknowledge the contributions of the more than 20 PhD students whom the author has mentored but who have always been considered as colleagues, and an even larger number of post-doctoral trainees, research assistants, and associates who have shared in the many hurdles and joys of the many discoveries that have been made. Many of their names are listed in the references to this article where there is shared authorship.
Support for the research has come from various sponsors, ranging from the National and Medical Research Councils of Canada, the National Institute of Dental and Craniofacial Research in the USA, and from numerous industrial supporters. Support from several of these sponsors and in particular from the late Mr. Ernie Sandler, President of IDE Interstate, and Mr. Mitchell Goldberg, President of Ortek Therapeutics, made it possible to weather the massive S. mutans mania that encompassed the dental research community for a good part of this period. Ironically, the author has rather enjoyed going against the conventional wisdom, since he has always believed that challenge and even adversity can be among the various joys of life.
The considerable assistance by Ms. Pat Calia in the preparation of this manuscript and by Dr. Mark Wolff in the preparation of the figures and valuable comments is most gratefully appreciated.