- Email: [email protected]
pH changes during simultaneous metabolism of urea and carbohydrate by human salivary bacteria in vitro
Archs oral Bid. Printed in Great
Vol. 33, Britain.
No. 8, pp. 579-587, All rights reserved
1988 Copyright
0
0003-9969/88 S3.00 + 0.00 1988 Pergamon Fkss plc
pH CHANGES DURING SIMULTANEOUS METABOLISM OF UREA AND CARBOHYDRATE BY HUMAN SALIVARY BACTERIA IN VITRO C. H. SISSONSand T. W. CUTRE~S Dental Research Unit, Medical Research Council of New Zealand, P.O. Box 27007, Wellington, New Zealand (Accepted 3 February 1988)
Summary-The effect of the wide natural variation in oral ureolysis rates on the pH changes resulting from simultaneous metabolism of 25 mM urea and 2.8 mM glucose in salivary-sediment bacteria were investigated. The pH curves were complex, and included distinctive plateaux indicative of balanced acid and base production. These neutralization plateaux occurred at different pHs, which were a function (r* = 0.98) of the ureolytic rate as measured by the log of the initial pH-change rate in the urea-only reaction. In the simplest case, the pH curve was characterized by a rise or fall to the neutralization plateau, a variable period of time at the plateau (up to 1 h), then a pH rise. The pattern of pH changes induced by glucose alone varied between different sediments: in some cases, the pH decreased smoothly to an end-point; in others, the curve was more complex, and these features became superimposed on the urea/glucose curve. The rate of ureolytic ammonia release was almost constant and unaffected by simultaneous carbohydrate metabolism. Concomitant metabolism of endogenous carbohydrate present in sediments prepared l-2 h following a meal was of sufficient magnitude to affect ureolytic pH curves. If the ureolytic activity was high, this et&t was negligible; if it was low, metabolism of the endogenous carbohydrates could completely suppress the ureolytic pH rise. Soluble salivary components had little effect on ureolvsis but DH chances were modified by buffering, and the presence of urea, ammonia, N-catabolic anb acidogkic subsGates in the saliva. INTRODUCTION
widely as acid is generated from periodically available carbohydrate in the presence of continuously generated alkali, and as saliva properties and flow rate vary. The processes causing acid fluctuations in plaque pH have been well studied (Geddes, 1984), but those controlling the return to the resting pH (i.e. the rise portion of the Stephan curve) are much less understood. Alkali formation (Biswas and Kleinberg, 1971; Kleinberg et al., 1979, 1982)., removal of oral carbohydrate (Lagerlof, Dawes and Dawes, 1984), further metabolism of organic acids, and alkaline salivary buffers (Jenkins, 1979) all contribute to the pH rise but their relative importance is unclear (Abelson and Mandel, 198 1). Ammonia generation to give an alkaline plaque may be a key factor in dental caries (Kleinberg et af., 1982; Peterson, Woodhead and Crall, 1985; Minah, McEnery and Flores, 1986), in the formation of dental calculus (Regolati, 1971; Epstein, Mandel and Scott, 1980) and in triggering active periodontitis (Aleo, Padh and !;ubramonisam, 1984; Helgeland, 1985). The major ammonia-generating substrate is probably urea, with contributions also from arginine and proline peptides (Kleinberg et al., 1979, 1982; Curtis and Kemp, 1984). Urea, although present at about I-IOmM in mixed normal saliva, is secreted at concentrations up to 25 mM in parotid saliva (Kopstein and Wrong, 1977), and is reported to be 55 mM in uninflamed gingival crevicular fluid (Golub, Borden and Kleinberg, 1971). Salivary urea levels are elevated in renal disease (Peterson et al., 1985). Investigation of ureolytic pH effects may The pH of dental plaque fluctuates
579
also improve the efficacy of a recently developed caries-protective mouthrinse through which an ureadependent pH-rise causes the deposition of calciumphosphate minerals into plaque (Pearce, 1981, 1982). Net mineral deposition in vivo varies lOO-fold (Pearce, 1984); several factors suggest that the interaction between metabolism of urea and low carbohydrate levels may be important in this process. First, oral ureolytic pH changes in the salivary sediment from different persons differ by at least lo-fold (Sissons and Cutress, 1987). The balance between ureolysis and acidogenesis from low levels of carbohydrate might change greatly with this large variation in ureolysis rate. Second, the mouthrinse is usually applied within 4 h of a meal-time when oral pH is still depressed (Kleinberg and Jenkins, 1964) and influencing ureolytic pH changes. Third, although plaque growth is mostly carbohydrate-limited (van der Hoeven et al., 1984), preliminary experiments have shown that in washed and unpre-incubated bacteria from salivary sediment, prepared l-2 h after meals, the endogenous carbohydrate often markedly affected urea-dependent pH changes. Saliva affects oral pH by buffering, dilution and clearance effects, and also by contributing acid- and base-generating substrates. Potentially all these factors greatly affect ureolytic pH changes. Simultaneous metabolism of urea with carbohydrate raises the pH minimum in the Stephan curve (Stephan, 1943), but pH curves, including long periods of constant pH, reported in in-v&o studies, have yet to be fully explained (Jenkins and Wright, 1951; Biswas, 1982). Our object was to investigate the pH changes resulting from concomitant metabolism of urea and low levels of carbohydrate by salivary bacteria with
580
C. H. Srsso~s and T. W. Cu~a~ss Saliva
I
Centrifuge
1250
g
,
25rntn,
0°C
/ Pellet
(Sediment)
Wash
Supernatant
2x Hz0
Resuspend
(centrrfugel,
in H20.
1 Portion for assay
_
Sediment
CHO-depleted
I
+Urea
/I\
+Urea
. . . . .. .. . . . .. .. . . Pre-Incubate 35°C 60min, . . .. .. Wash 1 x H,O 1centrifuge), . . .. .. Resuspend in H,O. . . . . .. .. .. -. . . sediment .. .. . . . . . . . . . .... . . * *. . *. .. .. .* i i i
with
t Urea +Glc
CliO.~~~~~~~~~~~~~~
\ t Glc
+Supernatant + Urea -Urea
Reaction Urea t CHO fsed.)
Urea-only contrai
Urea/Gtc reaction
Glc-only control
Urea-S/N reactions
S/N Controls
Fig. 1. Flow-chart of salivary sediment and supematant preparation and analysis. Sediment was present in reactions at 16.7 per cent (wet weight/v), urea at 25 mM and glucose at 2.8 mM. Dotted lines indicate experiments investigating salivary supematant effects on ureolysis, ammonia release and pH changes. Abbreviations: Glc = glucose; CHO = carbohydrate; CHO(sed.) = endogenous carbohydrate, that is carbohydrate existing in the sediment at the time of preparation; S/N = salivary supematant (33 per cent, v/v).
widely different rates of ureolysis and, in particular, to examine the sensitivity of the ureolytic pH rise to variation in ureolysis rate when carbohydrate is co-metabolized. Carbohydrate levels were chosen to mimic those that commonly occur after the rapid clearance of soluble carbohydrate from the oral cavity (Singer and Kleinberg, 1983; Sreebny, Chatterjee and Kleinberg, 1985), and included carbohydrates naturally present in salivary sediments prepared l-2 h after a meal. The influence of soluble salivary compounds on urea metabolism, ureolytic release was also pH changes, and ammonia examined. MATERIALS
AND METHODS
Experimental procedures
Changes in pH were determined in salivary sediments, depleted of endogenous carbohydrate (by pre-incubation), during the simultaneous metabolism of added urea (25 mM) and glucose (2.8 mM), and during metabolism of urea only, and glucose only. In sediments prepared at 0°C and not pre-incubated, pH changes generated by simultaneous metabolism of urea and the endogenous carbohydrate were compared with those generated by urea only. Ammonia release was measured in urea-only incubations and, in
some experiments, also in carbohydrate-containing reactions. Salivary supernatant was analysed for base-generating substrates, effects on ureolytic pH changes, and direct effects on ureolysis. The general procedure and the origin of salivary fractions is summarized in Fig. 1. Salivary sediment and supernatant preparation
To obtain sediment preparations with a range of ureolysis rates, samples were obtained from individuals who differed up to IO-fold in the ureolytic activity of their oral floras, as previously determined (Sissons and Cutress, 1987). Samples from individuals with similar ureolytic activities were pooled in experiments requiring large volumes of sediment. Saliva was collected, whilst chewing chicle gum, in ice-cooled tubes and then centrifuged at 125Og for 15 min at 0°C. The resuspended pellet (salivary sediment) was washed twice with water by centrifugation, at O”C, and finally resuspended in ice-cold water to 33 per cent (wet weight/vol). A portion (containing endogenous carbohydrate) was removed for analysis of urea-induced pH changes and, in some experiments, of ureolytic ammonia release. Endogenous carbohydrate in the remaining sediment was metabolized by incubation at 16.7 per cent (w/v) in water for 60min at 35°C. The pre-incubated sediment
pH changes in urea and glucose metabolism
/
I
I
30 Time
60 (min)
581
I
90
Fig. 2. The pH changes in a moderate ureolytic activity salivary sediment during metabolism of 25 mM urea only (O--O), 2.8 mM glucose only (A-A) and, urea and glucose (0-O). A second pH curve for simultaneous urea and glucose metabolism in a sediment prepared from the same individual on a different day is also shown (O-IJ).
preparation was centrifuged, resuspended in water, recentrifuged and aga.in resuspended in ice-cold water to 33 per cent (w/v), and then stored on ice. For experiments on salivary supernatant effects, ammonia and urea were measured in 50 p 1samples of supernatant added to 1 ml of 7 per cent trichloroacetic acid (in triplicate). Sediment preparation was modified in these experiments by resuspending the sediment at 50 per cent (w/v). Effects of premetabolizing salivary supernatant components were determined by preparing sediment and supernatant at room temperature to allow metabolism to proceed, resuspending the sediment in ice-cold water to 50 per cent (w/v), followed by storage at 0°C. Reaction conditions and analytical techniques
Eight hundred microlitres of resuspended salivary sediment (33 per cent w/v) were incubated at 35°C for I min and 800~1 of one of the following substrates (warmed to 35°C) added: 50mM urea, 5.4mM glucose, or 50 mM urea--5.4 mM glucose. This procedure was modified when studying the effects of salivary supernatant on the sediment system, where reaction mixtures comprised 500~1 of sediment (50 per cent w/v), 500 ~1 of salivary supernatant, and 500 ~1 of 75mM urea (or 500~1 of water as appropriate). All reactions were incubated at 35°C with stirring, and pH changes measured with a Radiometer (Copenhagen, Denmark) GK 2321C electrode and Corning (Halstead, England) Mode1 610 pH meter. Urea metabolism was measured by periodic sampling of the reaction mixture, 50 ~1 in triplicate, which was added to 1 ml of 7 per cent (w/v) trichloroacetic acid for subsequent ammonia and urea analysis. Ammonia was measured by the Berthelot indophenol reaction, and urea by diacetylmonoximethiosemicarbazide reaction (Sissons, Cutress and Pearce, 1985). Rates of pH-change (abbreviated to
ApH-rate and measured in pH units x lo3 per min) were extracted from linear regions of the pH curves generated. The initial ApH-rate obtained from pH curves of urea metabolism in the absence of carbohydrate was also used as an estimate of ureolysis rate (Sissons and Cutress, 1987). Values for pH were transformed, where appropriate, into PM concentrations of: [H+] = antilog (0 - pH), and [OH-] = antilog(pH - pKw). At 35°C pKw = 13.68 (Albert and Serjeant, 1971). RESULTS pH response to co-metabolism of urea and glucose
In the absence of carbohydrate (i.e. in preincubated sediments), urea metabolism gave typical pH-rise curves (control curves in Figs 2, 3A, and 6) as previously described (Sissons and Cutress, 1987). The simplest pH curve resulting from cometabolism of urea and glucose was most clearly seen in sediments with moderate ureolytic activity (Fig. 2). There was, generally, in sequence an initial decrease or increase in pH, a neutralization plateau, and a final pH rise. Sediment from the same subject (Fig. 2) but collected on a different day, showed a similar pattern and plateau pH (pH 6.8). Processes underlying the urea/glucose response
In a sediment with a 4-fold higher ureolytic activity than shown in Fig. 2, simultaneous metabolism of urea and glucose produced a high neutralization plateau at pH 7.9 (Fig. 3A). However the pH curve was more complex, with the pH rise occurring in phases of constant rate until the plateau was reached. These constant-rate periods mainly coincided with changes in the glucose-only pH curve (Fig. 3A). Changes in rate were observed in both the urea-glucose [OH-] and glucose-induced [H ‘1 curve
582
C.
H.
%SONS
and T. W. Curarzss
6
8
6
-I-
i-
90
60
30 Time
30
(minl
Ttme
10
20
30 Time
40
50
I
/
60
so
Lmlnl
60
(mini
Fig. 3. The pH changes in a high ureolytic-activity salivary sediment (pooled from 3 people) during metabolism of 25 mM urea and 2.8 mM glucose. (A) pH changes with; urea only (a---•), glucose only (A---&, urea and glucose (O----O), urea and endogenous carbohydrate (V-V). In the urea-glucose curve, segments approximating straight lines are marked as regions 14 and the corresponding regions also marked on the glucose-only curve. (B) [OH-] changes with; urea only (O-O), both urea and glucose (O----O), and the [H+] concentration change with glucose only (A-A). (C) Ureolytic ammonia release with; urea only (O-O), urea and glucose (A-A), urea and endogenous carbohydrate (O-0).
(Fig. 3B), also indicating that periods of constant rate occurred in those curves. The net concentration changes for [OH-] in the urea-only reaction, and [H+] in the glucose-only reaction, had no direct relationship to [OH-] changes in the urea-glucose reaction-except for the super-
imposed rate fluctuations (Fig. 3B). In fact, although co-metabolism of urea and glucose gave a large pH rise (Fig. 3A), the rate of [H+] increase in the glucose-only reaction was twice the maximal rate of [OH-] increase in the urea-only reaction, Between 24 and 29 min (the pH plateau in the urea-glucose
pH changes in urea and glucose metabolism
metabolizing urea only (Fig. 4). The initial rate of pH change measures the ureolysis rate (Sissons and
I
I
&tress,
I
1987).
If ureolytic activity in the sediment was very low, the pH continuously decreased during simultaneous metabolism of glucose and urea (Fig. 5). In Fig. 5 the theoretical plateau pH, from the relationship derived in Fig. 4, is pH 5.0.
8.5-
I, 90s 2 2
583
Effect of variable sediment glucose-induced responses on the wealglucose pH curve
75-
70-
2bh2
18
2.4
Log ureolysls
rate (A pH, x 103/min)
Fig. 4. Relationship between the pH of the urea/glucose neutralization plateau and ureolysis rate measured as the initial pH-change rate in the urea only reaction (ApHi). The regression equation (r*=O.98) was: plateau pH = 3.34 f 0.:!9 + (1.84 &-0.13) l log(ApHi *IO’).
reaction) the ratio of the rate of change in [H+] to [OH-] in the separate reactions was 5.5. After 30 min the pH rose to reach a final plateau. Metabolism of ghlcose (2.8 mM) had almost no effect on ureolytic ammonia release (Fig. 3C). The ammonia production rate was constant for about 50min, which included the different constant rate regions of the urea/glucose pH curve (Fig. 3A). Effect of variation in ureolysis rate on the neutralization plateau pH
The neutralization pH plateaux induced by simultaneous metabolism of 25 mM urea and 2.8 mM glucose were closely associated (r* = 0.98) with the initial rate of change in pH seen in sediments
Glucose-induced pH responses varied considerably between sediments, and these differences were reflected in the urea/glucose pH curve (Figs 2 and 3). A further type of response was also obtained (Fig. 6), in which the pH curve reached a transient low of pH 4.7 and then rose to a higher final plateau of pH 5.0. These changes were again mirrored in the urea/glucose curve where the pH rose to a plateau at pH 8.4 initially and then rose to a new plateau at a time approximately corresponding to the rise in the glucose-only curve (Fig. 6). Thus much of the complexity found in urea/glucose pH curves was derived from a changing pH response to glucose metabolism. EApcts of endogenous sediment carbohydrate
Endogenous carbohydrate (existing in the sediment before incubation) had negligible effects on ureolytic pH changes in active sediments. For example, in the sediment shown in Fig. 3A, it reduced the pH increased during the initial 5 min by I5 per cent, but subsequently the pH curves were 0.05 pH units apart and parallel. Ammonia release was unaffected (Fig. 3B). In another active sediment (Fig. 6), metabolism of endogenous carbohydrate generated only a short plateau at pH 8.7. The small effect on ureolytic pH changes generated by simultaneous metabolism of endogenous carbohydrate in high-activity ureolytic reactions was in contrast to a major effect in sediments having a low
7
Time
I
I
60
90
(min)
Fig. 5. Urea-dependent pH changes in a low ureolytic-activity sediment with; urea only (a----•), and glucose (O-O),
pH
and urea and endogenous carbohydrate (V-V).
urea
584
C. H. Srsso~s and T. W. Cum
incubated sediment gave similar results (data not shown). If saliva was fractionated into sediment and supernatant at room temperature to allow depletion of metabolizable substrates, the supematant had no effect on ureolytic ammonia release (Fig. 9). DISCUSSION
Ureolysis by salivary bacterial systems in vitro in the presence of low levels of carbohydrate generates complex pH curves that usually show distinctive plateaux resulting from neutralization of acid and base. From our results, in which the ureolysis rate varied, and those reported by Biswas (1982), where the acidogenesis rate varied, it is possible to consider the underlying pH pattern obtained as an approach to, and departure from, the neutralization pH. However, this simple pattern can be further complicated by changing rates of acidogenesis from carbohydrate. Analysis of pH curves created by simultaneous metabolism of urea and glucose in vitro
6b
3b Time
9b
(min)
Fig. 6. The pH changes in a high ureolytic-activity sediment which showed a Stephan curve-like response to 2.8 mM glucose only (A--A); 25 mM urea only (-), urea and glucose (O-O), urea and endogenous carbohydrate (V-V).
ureolytic activity. In extreme cases, simultaneous metabolism of urea and endogenous carbohydrate (as well as glucose) could suppress the ureolytic pH rise for long periods of time, or even cause a pH decrease
(Fig. 5). Eflects of soluble salivary components
Addition of salivary supernatant at 33 per cent (v/v) to a carbohydrate-depleted sediment generated an irregular curve with periods of constant pH change and a plateau, characteristic of co-metabolism of urea and carbohydrate (Fig. 7; compare with Fig. 3A). Supernatant components also raised the initial pH and lowered both the initial rate of pH increase and the final pH*ffects mainly caused by added buffers because ammonia release was only increased by 6 per cent (2SmM) over 90min (Fig. 8A). Reactions involving salivary sediment in the absence of urea, with and without supematant, showed that the supernatant caused an initial rapid increase in ammonia release, which slowed to a lower constant rate after 30min (Fig. 8B). The initial, faster component of the ammonia-release curve (approx. 1.8 mM NH, released over 30 min) probably derived from the 1.7mM urea-N present in this supernatant. The slow component came from at least two sources: about 15 per cent was ammonia released from the washed sediment (estimated from the control reaction, Fig. 8B); the remainder was probably generated by metabolism of soluble salivary components, accounting for the small supernatant-induced increase in ammonia release during ureolysis (Fig. 8A). Pre-
A neutralization plateau in the pH curve represents a steady-state constant ratio of acid and base production with no pH change. At any ratio of acid to base production there will be a pH at which they wiIl neutralize each other, the neutralization pH. Examples are the pH plateaux shown in Figs 2, 3, 6 and 7. Plateaux caused by cell buffers are ruled out by their absence in urea-only controls. On initiation of metabolism the pH moves towards its neutralization value. Whether the pH rises or falls depends on whether the ratio of acid/base production rates leads to a neutralization pH above or below the starting pH. When the steady-state ratio of acid to base generation changes, the plateau is terminated and the pH changes. In our investigation, the low quantity of carbohydrate present was consumed before the 25 mM urea; hence, after a plateau, the pH
9-
Iae
-% Time
ImInI
Fig. 7. Effect of salivary supematant on ureolvtic OH changes in a carbohydrate-depleted sediment (pooled from 3 people); with (O-O) and without (o-0) added salivary supematant (33 per cent v/v).
pH changes in urea and glucose metabolism
(A)
I
1
,
30
60
so
Time (mid
I
/ (B1
585
uted the final pH rise (after 4 h) to a rising rate of ammonia production, although their data for NH, and lactate production indicated that it was caused by a falling rate of lactate production. In the most extensive study of simultaneous urea/glucose metabolism, Biswas (1982) described, for the salivary sediment-supernatant system, both pH changes and net acid or base production at pH 7.0 in a pH-stat, as generated by 28 mM urea and varying high glucose concentrations (up to 15 per cent). When the pH was free to change, concomitant metabolism of urea with all the glucose concentrations tested gave a pH decrease from about pH 7.6 to a plateau-which decreased from pH 6.7 to 6.1 with increasing glucose concentration. The pH then rose after 30 min in 0.17 per cent glucose and decreased after 45 min in 0.5 per cent (and higher) glucose. No explanation of these biphasic pH curves was put forward. Relationship of the neutralization pH to the ureolysis rate
The pH of the neutralization plateau was almost entirely explained by the ureolysis rate (r2 = 0.98). This relationship could be described by an equation (Fig. 4) of the form: plateau pH = pK, + K, log(dNH,/dt)
L-----------J
60
30
Time
90
(mm)
Fig. 8. Effects of salivary supernatant on ammonia release. (A) Ureolytic ammonia release in a portion of the sediment shown in Fig. 7 but prepared at OT, incubated with (0-O) and without (a----0) added salivary supernatant (33 per cent). (Em)Ammonia release in the absence of added urea for the sediment shown in (A), with (0-O) and without (A---L,) added supematant, and ammonia changes in the absence of supematant in a sediment which had been pre-incubated for 60min (o-0).
increased,
and this resulted
where K, and K, are constants. If the cell concentration does not vary greatly, the ureolysis rate (dNI-I,/dt) can be derived from the initial rate of change in pH generated by metabolism of urea only (these are highly correlated, r2 = 0.90; Sissons and Cutress, 1987). Different salivary sediments were used to obtain widely different ureolysis rates. Because the pH at which the neutralization plateau occurred was largely explained by the ureolysis rate, there was evidently little variation in the steady-state constant rate of acid production between sediments. This lack of variation in steady-state acidogenesis rate is distinct from the actual pattern of resulting pH change, which
1
in the typical curve (Fig.
2) with 2.8 mM glucose present. When the maximal acidogenesis rate outlasts base generation (Biswas, 1982), the pH falls. A change to long periods of constant pH followed by a pH rise (or fall), generated by simultaneous metabolism of urea and glucose in uitro, have been described befclre, but the explanation of these patterns has been rather limited. Stephan (1943) examined dispersed plaque with varying urea concentrations and 6 per cent glucose and found that urea raised the minimum pH of the Stephan curve. Jenkins and Wright (1951) reported results for whole saliva that resemble our ‘basic pH pattern (Fig. 2). They attributed the shape of the pH curve to the gradual overwhelming of the lactic acid being produced by the accumulated buffering capacity of the ammonia also produced-which generated a pH plateau. However, at pH 7.0 (their plateau), neither ammonia nor lactate have significant bulfering capacity. They also attrib-
Time (mini
Fig. 9. Ammonia release in a sediment-supematant system, different from that shown in Fig. 8, which had been prepared at room temperature to deplete metabolizable substrates; plus (0-O) and minus (0-O) supernatant (33 per cent v/v).
586
C. H. S~sso~sand T. W. CUTRESS
could vary significantly between sediments. It is also possible that there was variation in acidogenesis rate but that it was closely (and presumably inversely) correlated with the ureolysis rate. It is possible to predict (from the regression equation) circumstances in which neutralization plateaux would not occur in the pH curve. The neutralization pH must be reached, followed by a significant period of steady-state neutralization. This may not occur when buffers present (Sissons and Cutress, 1987) sufficiently slow the approach to the plateau pH, and the initial pH is distant from the neutralization pH. The duration of maximal steady-state acid or base production is limited and may be over before the neutralization pH is reached. The reactions shown in Fig. 5, where the calculated neutralization pH was pH 5.0, support this prediction. The plateaux pH found after analysing sediment from the same individual, prepared on two different occasions, were identical. Further investigation may show that the neutralization plateau pH is a stable index of an individual’s ureolytic activity. Processes underlying the urea/glucose pH response
The relationship of the pH curves to underlying processes is complex and not readily susceptible to analysis. A neutralization plateau pH is related to the ratio of [OH+] to [H+] actually present by the equation (derived from the dissociation constant of water): plateau pH = 0.5 pK, + O.S*log[OH-]/[H+]. The rates of OH- and H+ generation are such At neutrality maintain FiH 20.5pK = 6.84 %5’C~~~ rates of net OHand H+ pro&ction should be approximately equal (which was confirmed in the study, see Fig. 2). As the plateau pH value changes from neutrality, the situation becomes more complex. The neutralization pH depends on the various processes, including hydration and ionization of NH,, CO,, and carboxylic to neutralization, net acids, which contribute acid/base change, generation of cell transmembrane pH gradients (Kashket and Kashket, 1985), and buffering by strong cellular buffers (Sissons and Cutress, 1987). Although urea hydrolysis to NH, is a relatively simple process, the interpretation of pH changes resulting from simple urea hydrolysis is still partial (Sissons and Cutress, 1987). Multiple acidic metabolites from glucose or more complex carbohydrates make analysis of acidogenesis much more uncertain and difficult. Thus, during a neutralization plateau the actual NH3 and R-COOH production ratio will be different from the ratio of net production of [OH-] and [H ‘1 as measured in separate urea- and glucose-only reactions, which in turn will be different from the [OH-]/[H+] ratio at the plateau pH. The apparently conflicting results obtained when investigating the relationship of urea/glucose pH changes are probably due to the complexity of the underlying metabolic processes. For example, in Fig. 3, the rate of net [H+] generation in the glucose-only reaction was twice the maximum rate of net [OH-] generation in the urea-only reaction. Despite this excess of acid over base production in the separate
reactions, in the combined urea/glucose reaction the pH increased rapidly to an alkaline pH plateau. Thus there is no direct relationship between changes in pH, [H+], or [OH-], resulting from simultaneous urea and glucose metabolism, and the urea-induced [OH-] changes and glucose-induced [H+] changes in the separate reactions. Biswas (1982) found other discrepancies, in particular between metabolism in the pH-stat and in experiments where the pH was free to change. For example, glucose-urea mixtures, giving net base production when measured in a pH 7.0 pH-stat, gave a substantial pH decrease to below neutrality if the pH was free to change. These discrepancies may partly be due to effects of cellular buffers and transmembrane pH gradients. Clearly it is wrong to extrapolate net acid- and base-generation rates from individual reactions to predict pH changes occurring during simultaneous metabolism of acid- and base-generating substratesin either a pH-stat or when the pH is free to move. InJuence of variable glucose metabolism on the pattern of ureolytic pH curves
The pattern of the urea/glucose and the urea/endogenous carbohydrate pH curves varied significantly between different sediments. Changes additional to the basic pattern occurred during nearconstant ammonia production, which is unaffected by simultaneous carbohydrate metabolism. Therefore, when further complexities were introduced into the urea-glucose pH curves, they mostly reflected changes in acidogenesis rates or other non-ureolytic processes, which could be seen in the glucose-only curves. Soluble salivary components
Soluble salivary components affecting pH include: urea and peptides, which are degradable to NH,, factors that modify carbohydrate catabolism, and buffers, especially HCO; and phosphates (Kleinberg et af., 1979, 1982; Jenkins, 1979; Abelson and Mandel, 1981). Salivary supernatant is frequently incorporated into in vitro salivary-sediment systems in order to model the in vivo situation (e.g. Sandham and Kleinberg, 1969; Biswas and Kleinberg, 1971; Kleinberg et al., 1979, 1982; Biswas, 1982). We found that salivary supernatant had no major effect on ureolysis rate. Base production from soluble salivary components except urea was slow and relatively minor compared to the ureolysis rate in a high ureolytic-activity sediment, but may be more significant in low ureolytic-activity sediments. Obviously the importance of all these alkali-generating sources need further investigation. In addition, pH buffers, and soluble compounds which were metabolized to acids, were evidently present in salivary supernatant and had a major effect on pH. ReIationships between ingested food and ureolytic pH changes during prophylactic plaque mineralization
The wide individual variation in ureolytic activity causes differences in net urea-dependent pH changes (Sissons and Cutress, 1987), which are amplified in the presence of metabolizable carbohydrate. The pH effects of endogenous carbohydrate (Kleinberg and Jenkins, 1964; Sandham and Kleinberg, 1969), pre-
pH changes in urea and glucose metabolism sumably
derived
from
meals,
ranged
from
slight
in bacterial systems with high ureolytic activity to periods of pH reduction in bacteria from individuals with low ureolytic activity and a high initial pH (Fig. 5). Endogenous carbohydrates may include bacteria1 polysaccharide stores (Sandham and Kleinberg, 1969) and slowly hydrolysed residual dietary and salivary polysaccharides. Further study is required to determine their exact relevance to plaque pH changes. If the pH rise is suppressed by concomitant metabolism of endogenous carbohydrate, mineral deposition in the prophylactic mouthrinse procedure (Pearce, 1982) must be greatly reduced or even absent. This effect may generate the lOO-fold range in net calcium-phosph.ate mineral deposited in vivo. Therefore to optimize use of this procedure, individuals with especially low ureolytic activity should have mouth rinse applications scheduled before meals and their pre-ureolysis oral pH raised by procedures such as the stimulation cf salivary flow, e.g. by chewing non-cariogenic gum Acknowledgement-WI:
gratefully thank Mrs Hancock for her expel-t technical assistance.
E.
M.
REFERENCES
Abelson D. C. and MandelI. D. (1981) The effect of saliva on plaque pH in oi~o. J. dent. Res. 60, 16341638. Albert A. and Serjeant E. J. (1971) The Derermination of lonizalion C0nstant.t. Chapman & Hall, London. Aleo J. J., Padh H. and Subramonisam A. (1984) Possible role of calcium in periodontal disease. J. Periodont. 55, 642-647.
Biswas S. D. (1982) Effect of urea on pH, ammonia, amino acids and lactic acid in the human salivary sediment system incubated with varying levels of glucose. Archs oral Biol. 21, 6834191.
Biswas S. D. and Kleinberg I. (1971) Effect of urea concentration on its utilization, on the pH and the formation of ammonia and carbon dioxide in a human salivary sediment system. Archs oral Biol. 16, 759-780. Curtis M. A. and Kemp C. W. (1984) Nitrogen metabolism in dental plaque. In: Cariology Today (Int. Congr. Zurich 1983) (Edited by Guggenheim B.) pp. 212-222. Karger, Basel. Epstein S. R., Mandczl I. and Scott I. W. (1980) Salivary composition and calculus formation in patients undergoing hemodialysis. J. Periodont. 51, 336338. Geddes D. A. M. (1984) Current view of plaque acidogenicity. In: Caricslogy Today (Int. Congr. Zurich 1983) (Edited by Guggenheim B.) pp. 199-204. Karger, Basel. Golub L. M., Borden S. M. and Kleinberg I. (1971) Urea content of gingival crevicular fluid and its relation to periodontal d&as: in humans. J. periodonf. Res. 6, 243-25 1. Helgeland K. (1985) pH and the effect of NH&I on human gingival fibroblasts. Stand. J. dent. Res. 93, 3945. Hoeven J. S. van der, Jong M. H., Rogers A. H. and Camp P. J. M. (1984) A ,:onceptual model for the co-existence of Streptococcus Spp and Actinomyces Spp in dental plaque. J. dent. Res. 63, 389-392. Jenkins G. N. (19791)Salivary effects on plaque pH. In:
587
Saliva and Dental Caries (Edited by Kleinberg I., Ellison S. A. and Mandel I. D.) Special Supplement of Microbiology Abstracts, 1979, pp. 307-322. Information Retrieval, Washington, DC Jenkins G. N. and Wright D. E. (1951) The role of ammonia in dental caries. II. Effect of ammonium salts and urea on salivary organisms. Br. dent. 1. 90, 117-130. Kashket S. and Kashket E. R. (1985) Dissipation of the proton motive force in oral streptococci by fluoride. Infecr. Immun. 40, 19-22. Kleinberg I. and Jenkins G. N. (1964) The pH of dental plaques in the different areas of the mouth before and after meals and their relationship to the pH and rate of flow of resting saliva. Archs oral Biol. 9, 493-516. Kleinberg I., Kanapka J. A., Chatterjee R., Craw D., D’Angelo N. and Sandham H. J. (1979) Metabolism of nitrogen by the oral mixed bacteria. In: Saliva and Dental Caries (Edited by Kleinberg I., Ellison S. A. and Mandel I. D.) Special Supplement of Microbiology Abstracts, pp. 357-377. Information Retrieval, Washington, D.C. Kleinberg I.. Jenkins G. N.. Chatteriee R. and Wiievweera L. (1982) The antimony’ pH electrode and its ;ole in the assessment and interpretation of dental plaque pH. J. denr. Res. 61, 1139-1147. Kopstein J. and Wrong 0. M. (1977) The origin and fate of salivary urea and ammonia in man. Clin. Sci. Mol. Med. 52, 9-17.
Lagerlof F., Dawes R. and Dawes C. (1984) Salivary clearance of sugar and its effects on pH change by Streptococcus mutans in an artificial mouth. J. dent. Res. 63, 12661270.
Minah G. E., McEnery C. and Flares J. A. (1986) Metabolic differences between saliva from caries-active and restoration-free children. Archs oral Biol. 31. 633438. Pearce E. I. F. (I 98 1) The artificial mineralizatibn of dental plaque. Front. oral Physiol. 3, 108-124. Pearce E. I. F. (1982) Effect of nlaaue mineralization on experimental denta; caries. Caries kes. 16, 46&47 1. Pearce E. I. F. (1984) Therapeutic modifications to the mineral ion composition of dental plaque. Caries Res. 18, 103-l 14.
Peterson S., Woodhead J. and Crall J. (1985) Caries resistance in children with chronic renal failure: plaque pH, salivary pH, and salivary composition. Pediat. Res. 19, 796799.
Regolati B. (1971) Ammonia and urea in oral pathophysiology-a literature review. Helv. odont. Acta VII, 139-146. Sandham H. J. and Kleinberg I. (1969) The effect of glucose concentration on the interrelation between glucose utilization, pH and carbohydrate storage in a salivary system. . Archs oral Biol. 14, 663418. Singer D. L. and Kleinbere I. (1983) The free amino acids in human dental plaquerAr;hs oral Biol. 28, 873-878. Sissons C. H. and Cutress T. W. (1987) In vitro ureadependent pH-changes by human salivary bacteria and dispersed artificial mouth, bacterial plaques. Archs oral Biol. 32, 181-189. Sissons C. H., Cutress T. W. and Pearce E. I. F. (1985) Kinetics and product stoichiometry of ureolysis by salivary bacteria from human mouths and artificial mouth plaques. Archs oral Biol. 30, 781-790. Sreebny L. M., Chatterjee R. and Kleinberg I. (1985) Clearance of glucose and sucrose from the saliva of human subjects. Archs oral Biol. 30, 269-274. Stephan R. M. (1943) The effect of urea in counteracting the influence of carbohydrates on the pH of dental plaques. J. dent. Res. 22, 63-71.
Vol. 33, Britain.
No. 8, pp. 579-587, All rights reserved
1988 Copyright
0
0003-9969/88 S3.00 + 0.00 1988 Pergamon Fkss plc
pH CHANGES DURING SIMULTANEOUS METABOLISM OF UREA AND CARBOHYDRATE BY HUMAN SALIVARY BACTERIA IN VITRO C. H. SISSONSand T. W. CUTRE~S Dental Research Unit, Medical Research Council of New Zealand, P.O. Box 27007, Wellington, New Zealand (Accepted 3 February 1988)
Summary-The effect of the wide natural variation in oral ureolysis rates on the pH changes resulting from simultaneous metabolism of 25 mM urea and 2.8 mM glucose in salivary-sediment bacteria were investigated. The pH curves were complex, and included distinctive plateaux indicative of balanced acid and base production. These neutralization plateaux occurred at different pHs, which were a function (r* = 0.98) of the ureolytic rate as measured by the log of the initial pH-change rate in the urea-only reaction. In the simplest case, the pH curve was characterized by a rise or fall to the neutralization plateau, a variable period of time at the plateau (up to 1 h), then a pH rise. The pattern of pH changes induced by glucose alone varied between different sediments: in some cases, the pH decreased smoothly to an end-point; in others, the curve was more complex, and these features became superimposed on the urea/glucose curve. The rate of ureolytic ammonia release was almost constant and unaffected by simultaneous carbohydrate metabolism. Concomitant metabolism of endogenous carbohydrate present in sediments prepared l-2 h following a meal was of sufficient magnitude to affect ureolytic pH curves. If the ureolytic activity was high, this et&t was negligible; if it was low, metabolism of the endogenous carbohydrates could completely suppress the ureolytic pH rise. Soluble salivary components had little effect on ureolvsis but DH chances were modified by buffering, and the presence of urea, ammonia, N-catabolic anb acidogkic subsGates in the saliva. INTRODUCTION
widely as acid is generated from periodically available carbohydrate in the presence of continuously generated alkali, and as saliva properties and flow rate vary. The processes causing acid fluctuations in plaque pH have been well studied (Geddes, 1984), but those controlling the return to the resting pH (i.e. the rise portion of the Stephan curve) are much less understood. Alkali formation (Biswas and Kleinberg, 1971; Kleinberg et al., 1979, 1982)., removal of oral carbohydrate (Lagerlof, Dawes and Dawes, 1984), further metabolism of organic acids, and alkaline salivary buffers (Jenkins, 1979) all contribute to the pH rise but their relative importance is unclear (Abelson and Mandel, 198 1). Ammonia generation to give an alkaline plaque may be a key factor in dental caries (Kleinberg et af., 1982; Peterson, Woodhead and Crall, 1985; Minah, McEnery and Flores, 1986), in the formation of dental calculus (Regolati, 1971; Epstein, Mandel and Scott, 1980) and in triggering active periodontitis (Aleo, Padh and !;ubramonisam, 1984; Helgeland, 1985). The major ammonia-generating substrate is probably urea, with contributions also from arginine and proline peptides (Kleinberg et al., 1979, 1982; Curtis and Kemp, 1984). Urea, although present at about I-IOmM in mixed normal saliva, is secreted at concentrations up to 25 mM in parotid saliva (Kopstein and Wrong, 1977), and is reported to be 55 mM in uninflamed gingival crevicular fluid (Golub, Borden and Kleinberg, 1971). Salivary urea levels are elevated in renal disease (Peterson et al., 1985). Investigation of ureolytic pH effects may The pH of dental plaque fluctuates
579
also improve the efficacy of a recently developed caries-protective mouthrinse through which an ureadependent pH-rise causes the deposition of calciumphosphate minerals into plaque (Pearce, 1981, 1982). Net mineral deposition in vivo varies lOO-fold (Pearce, 1984); several factors suggest that the interaction between metabolism of urea and low carbohydrate levels may be important in this process. First, oral ureolytic pH changes in the salivary sediment from different persons differ by at least lo-fold (Sissons and Cutress, 1987). The balance between ureolysis and acidogenesis from low levels of carbohydrate might change greatly with this large variation in ureolysis rate. Second, the mouthrinse is usually applied within 4 h of a meal-time when oral pH is still depressed (Kleinberg and Jenkins, 1964) and influencing ureolytic pH changes. Third, although plaque growth is mostly carbohydrate-limited (van der Hoeven et al., 1984), preliminary experiments have shown that in washed and unpre-incubated bacteria from salivary sediment, prepared l-2 h after meals, the endogenous carbohydrate often markedly affected urea-dependent pH changes. Saliva affects oral pH by buffering, dilution and clearance effects, and also by contributing acid- and base-generating substrates. Potentially all these factors greatly affect ureolytic pH changes. Simultaneous metabolism of urea with carbohydrate raises the pH minimum in the Stephan curve (Stephan, 1943), but pH curves, including long periods of constant pH, reported in in-v&o studies, have yet to be fully explained (Jenkins and Wright, 1951; Biswas, 1982). Our object was to investigate the pH changes resulting from concomitant metabolism of urea and low levels of carbohydrate by salivary bacteria with
580
C. H. Srsso~s and T. W. Cu~a~ss Saliva
I
Centrifuge
1250
g
,
25rntn,
0°C
/ Pellet
(Sediment)
Wash
Supernatant
2x Hz0
Resuspend
(centrrfugel,
in H20.
1 Portion for assay
_
Sediment
CHO-depleted
I
+Urea
/I\
+Urea
. . . . .. .. . . . .. .. . . Pre-Incubate 35°C 60min, . . .. .. Wash 1 x H,O 1centrifuge), . . .. .. Resuspend in H,O. . . . . .. .. .. -. . . sediment .. .. . . . . . . . . . .... . . * *. . *. .. .. .* i i i
with
t Urea +Glc
CliO.~~~~~~~~~~~~~~
\ t Glc
+Supernatant + Urea -Urea
Reaction Urea t CHO fsed.)
Urea-only contrai
Urea/Gtc reaction
Glc-only control
Urea-S/N reactions
S/N Controls
Fig. 1. Flow-chart of salivary sediment and supematant preparation and analysis. Sediment was present in reactions at 16.7 per cent (wet weight/v), urea at 25 mM and glucose at 2.8 mM. Dotted lines indicate experiments investigating salivary supematant effects on ureolysis, ammonia release and pH changes. Abbreviations: Glc = glucose; CHO = carbohydrate; CHO(sed.) = endogenous carbohydrate, that is carbohydrate existing in the sediment at the time of preparation; S/N = salivary supematant (33 per cent, v/v).
widely different rates of ureolysis and, in particular, to examine the sensitivity of the ureolytic pH rise to variation in ureolysis rate when carbohydrate is co-metabolized. Carbohydrate levels were chosen to mimic those that commonly occur after the rapid clearance of soluble carbohydrate from the oral cavity (Singer and Kleinberg, 1983; Sreebny, Chatterjee and Kleinberg, 1985), and included carbohydrates naturally present in salivary sediments prepared l-2 h after a meal. The influence of soluble salivary compounds on urea metabolism, ureolytic release was also pH changes, and ammonia examined. MATERIALS
AND METHODS
Experimental procedures
Changes in pH were determined in salivary sediments, depleted of endogenous carbohydrate (by pre-incubation), during the simultaneous metabolism of added urea (25 mM) and glucose (2.8 mM), and during metabolism of urea only, and glucose only. In sediments prepared at 0°C and not pre-incubated, pH changes generated by simultaneous metabolism of urea and the endogenous carbohydrate were compared with those generated by urea only. Ammonia release was measured in urea-only incubations and, in
some experiments, also in carbohydrate-containing reactions. Salivary supernatant was analysed for base-generating substrates, effects on ureolytic pH changes, and direct effects on ureolysis. The general procedure and the origin of salivary fractions is summarized in Fig. 1. Salivary sediment and supernatant preparation
To obtain sediment preparations with a range of ureolysis rates, samples were obtained from individuals who differed up to IO-fold in the ureolytic activity of their oral floras, as previously determined (Sissons and Cutress, 1987). Samples from individuals with similar ureolytic activities were pooled in experiments requiring large volumes of sediment. Saliva was collected, whilst chewing chicle gum, in ice-cooled tubes and then centrifuged at 125Og for 15 min at 0°C. The resuspended pellet (salivary sediment) was washed twice with water by centrifugation, at O”C, and finally resuspended in ice-cold water to 33 per cent (wet weight/vol). A portion (containing endogenous carbohydrate) was removed for analysis of urea-induced pH changes and, in some experiments, of ureolytic ammonia release. Endogenous carbohydrate in the remaining sediment was metabolized by incubation at 16.7 per cent (w/v) in water for 60min at 35°C. The pre-incubated sediment
pH changes in urea and glucose metabolism
/
I
I
30 Time
60 (min)
581
I
90
Fig. 2. The pH changes in a moderate ureolytic activity salivary sediment during metabolism of 25 mM urea only (O--O), 2.8 mM glucose only (A-A) and, urea and glucose (0-O). A second pH curve for simultaneous urea and glucose metabolism in a sediment prepared from the same individual on a different day is also shown (O-IJ).
preparation was centrifuged, resuspended in water, recentrifuged and aga.in resuspended in ice-cold water to 33 per cent (w/v), and then stored on ice. For experiments on salivary supernatant effects, ammonia and urea were measured in 50 p 1samples of supernatant added to 1 ml of 7 per cent trichloroacetic acid (in triplicate). Sediment preparation was modified in these experiments by resuspending the sediment at 50 per cent (w/v). Effects of premetabolizing salivary supernatant components were determined by preparing sediment and supernatant at room temperature to allow metabolism to proceed, resuspending the sediment in ice-cold water to 50 per cent (w/v), followed by storage at 0°C. Reaction conditions and analytical techniques
Eight hundred microlitres of resuspended salivary sediment (33 per cent w/v) were incubated at 35°C for I min and 800~1 of one of the following substrates (warmed to 35°C) added: 50mM urea, 5.4mM glucose, or 50 mM urea--5.4 mM glucose. This procedure was modified when studying the effects of salivary supernatant on the sediment system, where reaction mixtures comprised 500~1 of sediment (50 per cent w/v), 500 ~1 of salivary supernatant, and 500 ~1 of 75mM urea (or 500~1 of water as appropriate). All reactions were incubated at 35°C with stirring, and pH changes measured with a Radiometer (Copenhagen, Denmark) GK 2321C electrode and Corning (Halstead, England) Mode1 610 pH meter. Urea metabolism was measured by periodic sampling of the reaction mixture, 50 ~1 in triplicate, which was added to 1 ml of 7 per cent (w/v) trichloroacetic acid for subsequent ammonia and urea analysis. Ammonia was measured by the Berthelot indophenol reaction, and urea by diacetylmonoximethiosemicarbazide reaction (Sissons, Cutress and Pearce, 1985). Rates of pH-change (abbreviated to
ApH-rate and measured in pH units x lo3 per min) were extracted from linear regions of the pH curves generated. The initial ApH-rate obtained from pH curves of urea metabolism in the absence of carbohydrate was also used as an estimate of ureolysis rate (Sissons and Cutress, 1987). Values for pH were transformed, where appropriate, into PM concentrations of: [H+] = antilog (0 - pH), and [OH-] = antilog(pH - pKw). At 35°C pKw = 13.68 (Albert and Serjeant, 1971). RESULTS pH response to co-metabolism of urea and glucose
In the absence of carbohydrate (i.e. in preincubated sediments), urea metabolism gave typical pH-rise curves (control curves in Figs 2, 3A, and 6) as previously described (Sissons and Cutress, 1987). The simplest pH curve resulting from cometabolism of urea and glucose was most clearly seen in sediments with moderate ureolytic activity (Fig. 2). There was, generally, in sequence an initial decrease or increase in pH, a neutralization plateau, and a final pH rise. Sediment from the same subject (Fig. 2) but collected on a different day, showed a similar pattern and plateau pH (pH 6.8). Processes underlying the urea/glucose response
In a sediment with a 4-fold higher ureolytic activity than shown in Fig. 2, simultaneous metabolism of urea and glucose produced a high neutralization plateau at pH 7.9 (Fig. 3A). However the pH curve was more complex, with the pH rise occurring in phases of constant rate until the plateau was reached. These constant-rate periods mainly coincided with changes in the glucose-only pH curve (Fig. 3A). Changes in rate were observed in both the urea-glucose [OH-] and glucose-induced [H ‘1 curve
582
C.
H.
%SONS
and T. W. Curarzss
6
8
6
-I-
i-
90
60
30 Time
30
(minl
Ttme
10
20
30 Time
40
50
I
/
60
so
Lmlnl
60
(mini
Fig. 3. The pH changes in a high ureolytic-activity salivary sediment (pooled from 3 people) during metabolism of 25 mM urea and 2.8 mM glucose. (A) pH changes with; urea only (a---•), glucose only (A---&, urea and glucose (O----O), urea and endogenous carbohydrate (V-V). In the urea-glucose curve, segments approximating straight lines are marked as regions 14 and the corresponding regions also marked on the glucose-only curve. (B) [OH-] changes with; urea only (O-O), both urea and glucose (O----O), and the [H+] concentration change with glucose only (A-A). (C) Ureolytic ammonia release with; urea only (O-O), urea and glucose (A-A), urea and endogenous carbohydrate (O-0).
(Fig. 3B), also indicating that periods of constant rate occurred in those curves. The net concentration changes for [OH-] in the urea-only reaction, and [H+] in the glucose-only reaction, had no direct relationship to [OH-] changes in the urea-glucose reaction-except for the super-
imposed rate fluctuations (Fig. 3B). In fact, although co-metabolism of urea and glucose gave a large pH rise (Fig. 3A), the rate of [H+] increase in the glucose-only reaction was twice the maximal rate of [OH-] increase in the urea-only reaction, Between 24 and 29 min (the pH plateau in the urea-glucose
pH changes in urea and glucose metabolism
metabolizing urea only (Fig. 4). The initial rate of pH change measures the ureolysis rate (Sissons and
I
I
&tress,
I
1987).
If ureolytic activity in the sediment was very low, the pH continuously decreased during simultaneous metabolism of glucose and urea (Fig. 5). In Fig. 5 the theoretical plateau pH, from the relationship derived in Fig. 4, is pH 5.0.
8.5-
I, 90s 2 2
583
Effect of variable sediment glucose-induced responses on the wealglucose pH curve
75-
70-
2bh2
18
2.4
Log ureolysls
rate (A pH, x 103/min)
Fig. 4. Relationship between the pH of the urea/glucose neutralization plateau and ureolysis rate measured as the initial pH-change rate in the urea only reaction (ApHi). The regression equation (r*=O.98) was: plateau pH = 3.34 f 0.:!9 + (1.84 &-0.13) l log(ApHi *IO’).
reaction) the ratio of the rate of change in [H+] to [OH-] in the separate reactions was 5.5. After 30 min the pH rose to reach a final plateau. Metabolism of ghlcose (2.8 mM) had almost no effect on ureolytic ammonia release (Fig. 3C). The ammonia production rate was constant for about 50min, which included the different constant rate regions of the urea/glucose pH curve (Fig. 3A). Effect of variation in ureolysis rate on the neutralization plateau pH
The neutralization pH plateaux induced by simultaneous metabolism of 25 mM urea and 2.8 mM glucose were closely associated (r* = 0.98) with the initial rate of change in pH seen in sediments
Glucose-induced pH responses varied considerably between sediments, and these differences were reflected in the urea/glucose pH curve (Figs 2 and 3). A further type of response was also obtained (Fig. 6), in which the pH curve reached a transient low of pH 4.7 and then rose to a higher final plateau of pH 5.0. These changes were again mirrored in the urea/glucose curve where the pH rose to a plateau at pH 8.4 initially and then rose to a new plateau at a time approximately corresponding to the rise in the glucose-only curve (Fig. 6). Thus much of the complexity found in urea/glucose pH curves was derived from a changing pH response to glucose metabolism. EApcts of endogenous sediment carbohydrate
Endogenous carbohydrate (existing in the sediment before incubation) had negligible effects on ureolytic pH changes in active sediments. For example, in the sediment shown in Fig. 3A, it reduced the pH increased during the initial 5 min by I5 per cent, but subsequently the pH curves were 0.05 pH units apart and parallel. Ammonia release was unaffected (Fig. 3B). In another active sediment (Fig. 6), metabolism of endogenous carbohydrate generated only a short plateau at pH 8.7. The small effect on ureolytic pH changes generated by simultaneous metabolism of endogenous carbohydrate in high-activity ureolytic reactions was in contrast to a major effect in sediments having a low
7
Time
I
I
60
90
(min)
Fig. 5. Urea-dependent pH changes in a low ureolytic-activity sediment with; urea only (a----•), and glucose (O-O),
pH
and urea and endogenous carbohydrate (V-V).
urea
584
C. H. Srsso~s and T. W. Cum
incubated sediment gave similar results (data not shown). If saliva was fractionated into sediment and supernatant at room temperature to allow depletion of metabolizable substrates, the supematant had no effect on ureolytic ammonia release (Fig. 9). DISCUSSION
Ureolysis by salivary bacterial systems in vitro in the presence of low levels of carbohydrate generates complex pH curves that usually show distinctive plateaux resulting from neutralization of acid and base. From our results, in which the ureolysis rate varied, and those reported by Biswas (1982), where the acidogenesis rate varied, it is possible to consider the underlying pH pattern obtained as an approach to, and departure from, the neutralization pH. However, this simple pattern can be further complicated by changing rates of acidogenesis from carbohydrate. Analysis of pH curves created by simultaneous metabolism of urea and glucose in vitro
6b
3b Time
9b
(min)
Fig. 6. The pH changes in a high ureolytic-activity sediment which showed a Stephan curve-like response to 2.8 mM glucose only (A--A); 25 mM urea only (-), urea and glucose (O-O), urea and endogenous carbohydrate (V-V).
ureolytic activity. In extreme cases, simultaneous metabolism of urea and endogenous carbohydrate (as well as glucose) could suppress the ureolytic pH rise for long periods of time, or even cause a pH decrease
(Fig. 5). Eflects of soluble salivary components
Addition of salivary supernatant at 33 per cent (v/v) to a carbohydrate-depleted sediment generated an irregular curve with periods of constant pH change and a plateau, characteristic of co-metabolism of urea and carbohydrate (Fig. 7; compare with Fig. 3A). Supernatant components also raised the initial pH and lowered both the initial rate of pH increase and the final pH*ffects mainly caused by added buffers because ammonia release was only increased by 6 per cent (2SmM) over 90min (Fig. 8A). Reactions involving salivary sediment in the absence of urea, with and without supematant, showed that the supernatant caused an initial rapid increase in ammonia release, which slowed to a lower constant rate after 30min (Fig. 8B). The initial, faster component of the ammonia-release curve (approx. 1.8 mM NH, released over 30 min) probably derived from the 1.7mM urea-N present in this supernatant. The slow component came from at least two sources: about 15 per cent was ammonia released from the washed sediment (estimated from the control reaction, Fig. 8B); the remainder was probably generated by metabolism of soluble salivary components, accounting for the small supernatant-induced increase in ammonia release during ureolysis (Fig. 8A). Pre-
A neutralization plateau in the pH curve represents a steady-state constant ratio of acid and base production with no pH change. At any ratio of acid to base production there will be a pH at which they wiIl neutralize each other, the neutralization pH. Examples are the pH plateaux shown in Figs 2, 3, 6 and 7. Plateaux caused by cell buffers are ruled out by their absence in urea-only controls. On initiation of metabolism the pH moves towards its neutralization value. Whether the pH rises or falls depends on whether the ratio of acid/base production rates leads to a neutralization pH above or below the starting pH. When the steady-state ratio of acid to base generation changes, the plateau is terminated and the pH changes. In our investigation, the low quantity of carbohydrate present was consumed before the 25 mM urea; hence, after a plateau, the pH
9-
Iae
-% Time
ImInI
Fig. 7. Effect of salivary supematant on ureolvtic OH changes in a carbohydrate-depleted sediment (pooled from 3 people); with (O-O) and without (o-0) added salivary supematant (33 per cent v/v).
pH changes in urea and glucose metabolism
(A)
I
1
,
30
60
so
Time (mid
I
/ (B1
585
uted the final pH rise (after 4 h) to a rising rate of ammonia production, although their data for NH, and lactate production indicated that it was caused by a falling rate of lactate production. In the most extensive study of simultaneous urea/glucose metabolism, Biswas (1982) described, for the salivary sediment-supernatant system, both pH changes and net acid or base production at pH 7.0 in a pH-stat, as generated by 28 mM urea and varying high glucose concentrations (up to 15 per cent). When the pH was free to change, concomitant metabolism of urea with all the glucose concentrations tested gave a pH decrease from about pH 7.6 to a plateau-which decreased from pH 6.7 to 6.1 with increasing glucose concentration. The pH then rose after 30 min in 0.17 per cent glucose and decreased after 45 min in 0.5 per cent (and higher) glucose. No explanation of these biphasic pH curves was put forward. Relationship of the neutralization pH to the ureolysis rate
The pH of the neutralization plateau was almost entirely explained by the ureolysis rate (r2 = 0.98). This relationship could be described by an equation (Fig. 4) of the form: plateau pH = pK, + K, log(dNH,/dt)
L-----------J
60
30
Time
90
(mm)
Fig. 8. Effects of salivary supernatant on ammonia release. (A) Ureolytic ammonia release in a portion of the sediment shown in Fig. 7 but prepared at OT, incubated with (0-O) and without (a----0) added salivary supernatant (33 per cent). (Em)Ammonia release in the absence of added urea for the sediment shown in (A), with (0-O) and without (A---L,) added supematant, and ammonia changes in the absence of supematant in a sediment which had been pre-incubated for 60min (o-0).
increased,
and this resulted
where K, and K, are constants. If the cell concentration does not vary greatly, the ureolysis rate (dNI-I,/dt) can be derived from the initial rate of change in pH generated by metabolism of urea only (these are highly correlated, r2 = 0.90; Sissons and Cutress, 1987). Different salivary sediments were used to obtain widely different ureolysis rates. Because the pH at which the neutralization plateau occurred was largely explained by the ureolysis rate, there was evidently little variation in the steady-state constant rate of acid production between sediments. This lack of variation in steady-state acidogenesis rate is distinct from the actual pattern of resulting pH change, which
1
in the typical curve (Fig.
2) with 2.8 mM glucose present. When the maximal acidogenesis rate outlasts base generation (Biswas, 1982), the pH falls. A change to long periods of constant pH followed by a pH rise (or fall), generated by simultaneous metabolism of urea and glucose in uitro, have been described befclre, but the explanation of these patterns has been rather limited. Stephan (1943) examined dispersed plaque with varying urea concentrations and 6 per cent glucose and found that urea raised the minimum pH of the Stephan curve. Jenkins and Wright (1951) reported results for whole saliva that resemble our ‘basic pH pattern (Fig. 2). They attributed the shape of the pH curve to the gradual overwhelming of the lactic acid being produced by the accumulated buffering capacity of the ammonia also produced-which generated a pH plateau. However, at pH 7.0 (their plateau), neither ammonia nor lactate have significant bulfering capacity. They also attrib-
Time (mini
Fig. 9. Ammonia release in a sediment-supematant system, different from that shown in Fig. 8, which had been prepared at room temperature to deplete metabolizable substrates; plus (0-O) and minus (0-O) supernatant (33 per cent v/v).
586
C. H. S~sso~sand T. W. CUTRESS
could vary significantly between sediments. It is also possible that there was variation in acidogenesis rate but that it was closely (and presumably inversely) correlated with the ureolysis rate. It is possible to predict (from the regression equation) circumstances in which neutralization plateaux would not occur in the pH curve. The neutralization pH must be reached, followed by a significant period of steady-state neutralization. This may not occur when buffers present (Sissons and Cutress, 1987) sufficiently slow the approach to the plateau pH, and the initial pH is distant from the neutralization pH. The duration of maximal steady-state acid or base production is limited and may be over before the neutralization pH is reached. The reactions shown in Fig. 5, where the calculated neutralization pH was pH 5.0, support this prediction. The plateaux pH found after analysing sediment from the same individual, prepared on two different occasions, were identical. Further investigation may show that the neutralization plateau pH is a stable index of an individual’s ureolytic activity. Processes underlying the urea/glucose pH response
The relationship of the pH curves to underlying processes is complex and not readily susceptible to analysis. A neutralization plateau pH is related to the ratio of [OH+] to [H+] actually present by the equation (derived from the dissociation constant of water): plateau pH = 0.5 pK, + O.S*log[OH-]/[H+]. The rates of OH- and H+ generation are such At neutrality maintain FiH 20.5pK = 6.84 %5’C~~~ rates of net OHand H+ pro&ction should be approximately equal (which was confirmed in the study, see Fig. 2). As the plateau pH value changes from neutrality, the situation becomes more complex. The neutralization pH depends on the various processes, including hydration and ionization of NH,, CO,, and carboxylic to neutralization, net acids, which contribute acid/base change, generation of cell transmembrane pH gradients (Kashket and Kashket, 1985), and buffering by strong cellular buffers (Sissons and Cutress, 1987). Although urea hydrolysis to NH, is a relatively simple process, the interpretation of pH changes resulting from simple urea hydrolysis is still partial (Sissons and Cutress, 1987). Multiple acidic metabolites from glucose or more complex carbohydrates make analysis of acidogenesis much more uncertain and difficult. Thus, during a neutralization plateau the actual NH3 and R-COOH production ratio will be different from the ratio of net production of [OH-] and [H ‘1 as measured in separate urea- and glucose-only reactions, which in turn will be different from the [OH-]/[H+] ratio at the plateau pH. The apparently conflicting results obtained when investigating the relationship of urea/glucose pH changes are probably due to the complexity of the underlying metabolic processes. For example, in Fig. 3, the rate of net [H+] generation in the glucose-only reaction was twice the maximum rate of net [OH-] generation in the urea-only reaction. Despite this excess of acid over base production in the separate
reactions, in the combined urea/glucose reaction the pH increased rapidly to an alkaline pH plateau. Thus there is no direct relationship between changes in pH, [H+], or [OH-], resulting from simultaneous urea and glucose metabolism, and the urea-induced [OH-] changes and glucose-induced [H+] changes in the separate reactions. Biswas (1982) found other discrepancies, in particular between metabolism in the pH-stat and in experiments where the pH was free to change. For example, glucose-urea mixtures, giving net base production when measured in a pH 7.0 pH-stat, gave a substantial pH decrease to below neutrality if the pH was free to change. These discrepancies may partly be due to effects of cellular buffers and transmembrane pH gradients. Clearly it is wrong to extrapolate net acid- and base-generation rates from individual reactions to predict pH changes occurring during simultaneous metabolism of acid- and base-generating substratesin either a pH-stat or when the pH is free to move. InJuence of variable glucose metabolism on the pattern of ureolytic pH curves
The pattern of the urea/glucose and the urea/endogenous carbohydrate pH curves varied significantly between different sediments. Changes additional to the basic pattern occurred during nearconstant ammonia production, which is unaffected by simultaneous carbohydrate metabolism. Therefore, when further complexities were introduced into the urea-glucose pH curves, they mostly reflected changes in acidogenesis rates or other non-ureolytic processes, which could be seen in the glucose-only curves. Soluble salivary components
Soluble salivary components affecting pH include: urea and peptides, which are degradable to NH,, factors that modify carbohydrate catabolism, and buffers, especially HCO; and phosphates (Kleinberg et af., 1979, 1982; Jenkins, 1979; Abelson and Mandel, 1981). Salivary supernatant is frequently incorporated into in vitro salivary-sediment systems in order to model the in vivo situation (e.g. Sandham and Kleinberg, 1969; Biswas and Kleinberg, 1971; Kleinberg et al., 1979, 1982; Biswas, 1982). We found that salivary supernatant had no major effect on ureolysis rate. Base production from soluble salivary components except urea was slow and relatively minor compared to the ureolysis rate in a high ureolytic-activity sediment, but may be more significant in low ureolytic-activity sediments. Obviously the importance of all these alkali-generating sources need further investigation. In addition, pH buffers, and soluble compounds which were metabolized to acids, were evidently present in salivary supernatant and had a major effect on pH. ReIationships between ingested food and ureolytic pH changes during prophylactic plaque mineralization
The wide individual variation in ureolytic activity causes differences in net urea-dependent pH changes (Sissons and Cutress, 1987), which are amplified in the presence of metabolizable carbohydrate. The pH effects of endogenous carbohydrate (Kleinberg and Jenkins, 1964; Sandham and Kleinberg, 1969), pre-
pH changes in urea and glucose metabolism sumably
derived
from
meals,
ranged
from
slight
in bacterial systems with high ureolytic activity to periods of pH reduction in bacteria from individuals with low ureolytic activity and a high initial pH (Fig. 5). Endogenous carbohydrates may include bacteria1 polysaccharide stores (Sandham and Kleinberg, 1969) and slowly hydrolysed residual dietary and salivary polysaccharides. Further study is required to determine their exact relevance to plaque pH changes. If the pH rise is suppressed by concomitant metabolism of endogenous carbohydrate, mineral deposition in the prophylactic mouthrinse procedure (Pearce, 1982) must be greatly reduced or even absent. This effect may generate the lOO-fold range in net calcium-phosph.ate mineral deposited in vivo. Therefore to optimize use of this procedure, individuals with especially low ureolytic activity should have mouth rinse applications scheduled before meals and their pre-ureolysis oral pH raised by procedures such as the stimulation cf salivary flow, e.g. by chewing non-cariogenic gum Acknowledgement-WI:
gratefully thank Mrs Hancock for her expel-t technical assistance.
E.
M.
REFERENCES
Abelson D. C. and MandelI. D. (1981) The effect of saliva on plaque pH in oi~o. J. dent. Res. 60, 16341638. Albert A. and Serjeant E. J. (1971) The Derermination of lonizalion C0nstant.t. Chapman & Hall, London. Aleo J. J., Padh H. and Subramonisam A. (1984) Possible role of calcium in periodontal disease. J. Periodont. 55, 642-647.
Biswas S. D. (1982) Effect of urea on pH, ammonia, amino acids and lactic acid in the human salivary sediment system incubated with varying levels of glucose. Archs oral Biol. 21, 6834191.
Biswas S. D. and Kleinberg I. (1971) Effect of urea concentration on its utilization, on the pH and the formation of ammonia and carbon dioxide in a human salivary sediment system. Archs oral Biol. 16, 759-780. Curtis M. A. and Kemp C. W. (1984) Nitrogen metabolism in dental plaque. In: Cariology Today (Int. Congr. Zurich 1983) (Edited by Guggenheim B.) pp. 212-222. Karger, Basel. Epstein S. R., Mandczl I. and Scott I. W. (1980) Salivary composition and calculus formation in patients undergoing hemodialysis. J. Periodont. 51, 336338. Geddes D. A. M. (1984) Current view of plaque acidogenicity. In: Caricslogy Today (Int. Congr. Zurich 1983) (Edited by Guggenheim B.) pp. 199-204. Karger, Basel. Golub L. M., Borden S. M. and Kleinberg I. (1971) Urea content of gingival crevicular fluid and its relation to periodontal d&as: in humans. J. periodonf. Res. 6, 243-25 1. Helgeland K. (1985) pH and the effect of NH&I on human gingival fibroblasts. Stand. J. dent. Res. 93, 3945. Hoeven J. S. van der, Jong M. H., Rogers A. H. and Camp P. J. M. (1984) A ,:onceptual model for the co-existence of Streptococcus Spp and Actinomyces Spp in dental plaque. J. dent. Res. 63, 389-392. Jenkins G. N. (19791)Salivary effects on plaque pH. In:
587
Saliva and Dental Caries (Edited by Kleinberg I., Ellison S. A. and Mandel I. D.) Special Supplement of Microbiology Abstracts, 1979, pp. 307-322. Information Retrieval, Washington, DC Jenkins G. N. and Wright D. E. (1951) The role of ammonia in dental caries. II. Effect of ammonium salts and urea on salivary organisms. Br. dent. 1. 90, 117-130. Kashket S. and Kashket E. R. (1985) Dissipation of the proton motive force in oral streptococci by fluoride. Infecr. Immun. 40, 19-22. Kleinberg I. and Jenkins G. N. (1964) The pH of dental plaques in the different areas of the mouth before and after meals and their relationship to the pH and rate of flow of resting saliva. Archs oral Biol. 9, 493-516. Kleinberg I., Kanapka J. A., Chatterjee R., Craw D., D’Angelo N. and Sandham H. J. (1979) Metabolism of nitrogen by the oral mixed bacteria. In: Saliva and Dental Caries (Edited by Kleinberg I., Ellison S. A. and Mandel I. D.) Special Supplement of Microbiology Abstracts, pp. 357-377. Information Retrieval, Washington, D.C. Kleinberg I.. Jenkins G. N.. Chatteriee R. and Wiievweera L. (1982) The antimony’ pH electrode and its ;ole in the assessment and interpretation of dental plaque pH. J. denr. Res. 61, 1139-1147. Kopstein J. and Wrong 0. M. (1977) The origin and fate of salivary urea and ammonia in man. Clin. Sci. Mol. Med. 52, 9-17.
Lagerlof F., Dawes R. and Dawes C. (1984) Salivary clearance of sugar and its effects on pH change by Streptococcus mutans in an artificial mouth. J. dent. Res. 63, 12661270.
Minah G. E., McEnery C. and Flares J. A. (1986) Metabolic differences between saliva from caries-active and restoration-free children. Archs oral Biol. 31. 633438. Pearce E. I. F. (I 98 1) The artificial mineralizatibn of dental plaque. Front. oral Physiol. 3, 108-124. Pearce E. I. F. (1982) Effect of nlaaue mineralization on experimental denta; caries. Caries kes. 16, 46&47 1. Pearce E. I. F. (1984) Therapeutic modifications to the mineral ion composition of dental plaque. Caries Res. 18, 103-l 14.
Peterson S., Woodhead J. and Crall J. (1985) Caries resistance in children with chronic renal failure: plaque pH, salivary pH, and salivary composition. Pediat. Res. 19, 796799.
Regolati B. (1971) Ammonia and urea in oral pathophysiology-a literature review. Helv. odont. Acta VII, 139-146. Sandham H. J. and Kleinberg I. (1969) The effect of glucose concentration on the interrelation between glucose utilization, pH and carbohydrate storage in a salivary system. . Archs oral Biol. 14, 663418. Singer D. L. and Kleinbere I. (1983) The free amino acids in human dental plaquerAr;hs oral Biol. 28, 873-878. Sissons C. H. and Cutress T. W. (1987) In vitro ureadependent pH-changes by human salivary bacteria and dispersed artificial mouth, bacterial plaques. Archs oral Biol. 32, 181-189. Sissons C. H., Cutress T. W. and Pearce E. I. F. (1985) Kinetics and product stoichiometry of ureolysis by salivary bacteria from human mouths and artificial mouth plaques. Archs oral Biol. 30, 781-790. Sreebny L. M., Chatterjee R. and Kleinberg I. (1985) Clearance of glucose and sucrose from the saliva of human subjects. Archs oral Biol. 30, 269-274. Stephan R. M. (1943) The effect of urea in counteracting the influence of carbohydrates on the pH of dental plaques. J. dent. Res. 22, 63-71.