Binding of corroded ions to human saliva

Bindingof corrodedions to human saliva

H.J. Mueller Council on Dental Materials and Divisions of Biochemistry Chicago Avenue, Chicago, Ill 606 11, USA (Received 3 August 1984; revised 19 November 1984)

and Toxicology, American

Dental Association,

2 11, East

Employing equilibrium dialysis, the binding abilities of Cu. Al, Co and Cr ions from corroded Cu-AI and Co-Cr dental casting alloys towards human saliva and two of its gel chromatographic fractions were determined. Results indicate that both Cu and Co bind to human saliva i.e. 0.045 and 0.027 mg/mg protein, respectively.

Besides possessing

the largest binding ability, Cu also possessed the largest binding capacity. The saturation of Cu binding was not reached up to the limit of 0.35 mg protein/ml mg protein/ml.

employed in the tests, while Co reached full saturation at about 0.2

Chromium showed absolutely no binding to human saliva while Al ions did not pass through the

dialysis membranes. Compared to the binding with solutions that were synthetically salivary-type

made up to contain added

proteins, it is shown that the binding to human saliva is about 1 order of magnitude larger, at least

for Cu ions. Keywords:

Dental

materials,

metal,

chromatography,

saliva, protein,

The clearance rate of corrosion products from the mouth, that are in equilibrium with metallic components, provides a driving force for the continuation of the corrosion processes. Excretion, absorption and detoxication of these products are consequences likely to occur during passage through the alimentary track. If, however, the products are able to permeate the mucous substances lining the oral mucosa and translocate cell membranes and intracellular matter, then besides clearance via swallowing, absorption and detoxication clearance via the permeability of the oral mucosa becomes important. By these means events can take place which lead to undesirable biochemistries i.e. some metals increase the permeability of cell membranes by interacting electrostatically with their phospholipid layers by reducing surface negativity and changing conformation. Likewise, multivalent cations sensitize etthrocytes to the agglutinating action of antibodies by the attachment of proteins to the erthrocyte’s surfaces. Therefore, it can be shown that metals binding strongly to the salivary glycoproteins and mucopolysaccharides lining the exterior surfaces of the mucosa have tendencies which prevent these undesirable bioevents from occurring. The continual movement of these mucosubstances towards the back of the mouth leads to clearance and replenishment through swallowing. Yet in other situations metal-salivary protein interactions promote bioevents which lead to undesirable biochemistries. Some of these include the premature activation of enzymes by stabilization of their active sites, denaturation of proteins by the crosslinking of their essential side chains, and possibly also by the formation and attachment of plaque onto the pellicle of teeth”‘. The ability for selected corroded ions from dental alloys to bind to some salivary-type proteins in simulated @ 1985 146

Biomaterials

1985,

Vol 6 May

glycoprotein,

binding

saliva solutions indicates large differences in behaviour among some of the various metals3. However, the interaction of corroded ions with human saliva and with its fractionated components may be different from simulated saliva solutions containing added bulk proteins either individually or collectively and from sources not necessarily of human origin. Therefore this study was made to determine the binding of corroded ions to human saliva secretions and to compare these with the results previously obtained for the simulated solutions.

MATERIALS

AND

METHODS

Whole mixed human saliva was first collected without stimulation from one individual over a several hour period by having the person expectorate into a glass beaker surrounded by ice. The saliva sample was then centrifuged at 3000 rpm for 30 min. Upon decantation portions of the 200 ml supernatant sample were subjected to one of the following: i) a 2-fold concentration to 0.35 mg protein/ml in dialysis membranes by dehydration with polyethylene glycol solution; ii) protein precipitation by adding 20% of a 10% solution of trichloroacetic acid; and iii) about an 8fold dilution to 0.02 mg protein/ml by adding either the protein-free solution from above or by adding the Tris buffer solution as defined below. Protein determination was made by the Lowry assay method4. About 1 ml of the saliva concentrate was fractionated by gel chromatography using Sephacryl S-200* and a column+ having dimensions of 26 mm diam. by 730 mm length. A 50 mM Tris buffer, containing 50 mM NaCl and adjusted to pH 7 ‘Pharmacia, Piscataway, NJ 08854 ‘LKB, Gaithersburg, MD 20877 Butterworth

Et Co (Publishers)

Ltd. 0142-9812/85/030146-04$03.00

Binding of corroded ions: H.J. Mueller

Table 1

Binding of corroded ions to human

saliva (ppm)

Corroded

Corroded

0.35 mg/ml protein

I

Band II

protein

0.17 mg/ml protein

Band

concentration

Protein-free saliva

0.02

system

(0.02

mg/ml)

(0.02

cu Al

15.00 1.40

14.64 0

16.32 0

22.44 0

29.28 0

15.24 0

15.00 0

Nobilium co Cr

8.80 3.36

8.30 3.24

9.00 3.18

11.60 3.30

11.80 3.06

8.80 3.36

8.90 3.24

ion

mg/ml

mg/ml)

Aurorium

by HCI, was passed at about 1 ml/min in an ascending direction by a peristaltic pump. About 4.6 ml fractions were collected with a fraction collector system. The void volume of 145 ml of the column was determined by eluting blue dextran with a molecular weight of 2 X 1 06. Calibration of the column was done by eluting proteins of known molecular weight, which included ribonuclease-A (13 700), chymotrypsinogen-A (25 000), bovine serum albumin (67 000). and aldolase (158 000). The optical absorbance at 280 nm was analysed for each fraction with a dual beam grating spectrometer. Besides the centrifuged human saliva and the three solutions listed above, the two chromatographic bands were also submitted to the binding tests. Binding experiments were performed by enclosing the saliva samples in 1000 molecular weight cutoff dialysis membranes and dialysed against solutions containing ions of Cu, Al, Co, or Cr, for 24 h in a cold room utilizing a multiple dialysis unit*. These solutions were obtained by corroding dental alloys of Aurorium+ (71% Cu. 20%AI, 5% Ni, 3% Fe, and 1% Mn) and Nobilium* (63% Co, 28% Cr, and with Ni, Cu, and Fe) at 100 pa/cm’ for varying times up to 5 wk. The metallic ion concentrations of these solutions as well as the protein solutions inside the dialysis membranes were analysed by atomic absorption spectrophotometry.

RESULTS Figure 1 shows the chromatography profile by the elution of the 1 ml saliva concentrate from the Sephacryl S-200 column. Optical absorbance at 280 nm is plotted versus fraction number. One intense band (I) is obtained at an elution volume of V, = 153 ml, which is just slightly greater than the volume of the blue dextran void volume of V, = 145 ml, and a second band (II) of much broader profile and of lower intensities is also obtained much later in the elution. From a plot of the distribution coefficient, vE

-

vO

vT

-

vO

K ave= __-

tein concentration to the upper limit of concentration used, whereas Co binding reaches a plateau value of well below 0.35 mg protein/ml. The slopes of these lines correspond to the binding abilities which are 0.045 and 0.027 mg metal/mg protein for Cu and Co, respectively. ForCr( Table 1 and Figure 4) the binding to saliva supernatant was zero, while Al did not succumb to dialysis. The Cu and Co binding to the separate chromatographic fractions indicated similar trends at the 0.02 mg protein/ml concentration used (Table 7).

DISCUSSION The elution patterns obtained in this study by gel chromatography are similar to those that have been reported with human saliva supernatan?‘. The first intense high molecular weight band has been assayed and contains high concentrations of both protein and carbohydrate as hexosamine, while the patterns later in the elution are usually lower in intensity, broader in behaviour, and with any additional protein and carbohydrate peaks occurring not necessarily in the same positions. Figure 7 simply shows there are two major protein bands in the saliva collected for this study. The broadness and profile of the second band suggest that it may consist of several distinct protein fractions. By using linear salt gradient chromatography*“, fractions similar to fraction II of this study have been able to be separated into at least 3-4 distinct bands. Previous gel chromatography procedures with human saliva have almost exclusively used various types

0.6

f%

-

-‘53

m’

and with V, equal to the total volume of

the column, it is determined that band I corresponds to molecularweights of about 500 000 and greaterand band II corresponds to molecular weights of between 10 000 and 60 000. Table 7 presents the binding results for Cu, Al, Co, and Cr. In each instance the concentration of the corroded solution is presented as well as the concentration of the solutions inside the dialysis membranes after dialysis. For both Cu and Co appreciable binding occurred to the human saliva supernatant. Figures 2 and 3 plot concentration versus protein concentration of these two systems. Copper binding shows a continual linear trend with pro‘Pope. Menomonee Falls, WI 53051 tAurorium, Orlando, FL 32803 ‘Nobilium, American Gold, Albany, NY 12201

IO

20

30

40

50

60

70

80

90

100

110 120

Fraction no Figure 1 Elution profile of 1 ml human saliva supernatant containing 0.35 mg protein from a Sephacryl S-200 column having dimensions of 26 mm diam. by 730 mm long. The matrix solution of a 50 rnM Tris buffer with 50 mM NaCI and HCI to pH 7 was eluted in an ascending direction at about 1 ml/min. Each fraction corresponds to 4.6 ml. The blue dextran void volume is V, = 145 ml, while the total volume of the column V, = 390 ml. Band I corresponds to an elution volume V, = 153 ml, whereas band II covers a wide range.

Biomaterials

1985,

Vol 6 May

147

3ind~ng of corroded ions: H. J. Mueller

0.1

0.3

0.2 mg/ml

0.4

protein

Figure 2 The bindinS ability of corroded copper ions from Aurorium dentalcasting alloy to human saliva supernatant versus protein concentration in the saliva.

I

I

I

I

0.1

0.2

0.3

0.4

mgfml

1

protein

Figure 3 The binding ability of corroded cobalt ions from Nob~~iurn dental casting alloy to human saliva supernatant versus protein concentration in the saliva.

4t Ei d ci

3-

0 fi

”

0

2I

I

I

I

0.1

0.2

a3

mg/ml

I _ 0.4

protein

Figure 4 The binding ability of corroded chromium ions from Nobilium dental casting a/lay to human saliva supernatant versus protein concentration in the saliva.

of Sephadex’ (dextran) gels. Further separation and purification procedures of particular bands have utilized other gel filtration materials, like agarose, adsorption chromatography with hydroxyapatite, ion exchange chromatography, electrophoresis, and isoelectric focussing. With gel chromatography complete elution of all components should be complete with a volume equal to about the volume of the column to insure no interactional effects between sample species and gel beads. With Sephadex this is what is reported. With the results from ‘Pharmacia

148

Biomaterials

7985,

Vol6

Msy

this study using Sephacryl, which is another form of s dextran gel, a small amount of interaction must have occurred as indicated by the elution volumes. Since an upper limit of about 550 ml was needed to elute the last portions of the sample and with a column volume of 390 ml, a volume of about 160 ml in excess was needed to completely elute the sample. The essential difference between Sephadex and Sephacryl is the type of crosslinking occurring between the dextran molecules. The binding ability of corroded Cu from Aurorium to human saliva supematant is greater than the binding ability(as has been reported) of corroded Cu from Sybraloy amalgam to glycoprotein added solutions. These bindings correspond to 0.045 and 0.0065 mg Culmg protein, respectively. Two differences between these particular systems must be noted. Copper ions in solution generated from different sources may interact differently with protein and changes in the composition of the supporting electrolyte may also change interactional behaviour. Because of the similarity in chloride ion concentrations between both supporting electrolytes, it is reasonable to consider the diversity of species in human saliva responsible for the observed differences in binding. Various proteins, glycoproteins, carbohydrates and lipids all contained in human saliva may very well account for increased binding. Studies are needed to eliminate any possible effect from variations in supporting electrolyte. This can be overcome by dialysis of the human saliva samples before binding tests. The binding of Crfrom Nobilium in human saliva like the binding of Cr from Vitallium in glycoprotein added solutions3 are the same i.e. essentially zero. As has previously been suggested, the large amount of hydration occurring around Cr ions may hinder their attachment onto binding sites on the protein molecules. Corroded Al ions did not pass through the dialysis membranes possibly because of the very strong ligands that are formed in solution. These results were consistently obtained for all solutions submitted to dialysis. This same effect was noticed with corroded Sn ions from Sybraloy amalgam3. Discrepancies between binding of Cu and Co on fractionated versus unfractionated saliva (Table I) indicate whole saliva should be used to test the binding of metal ions. There is currently insufficient data to quantitate synergistic effects of combined fractions on ion binding. This study suggests that metals like Cu, Co, and Ni, which bind strongly to the organic salivary matter, have a better chance of being cleared from the mouth than metals like Cr which exhibit no binding to salivary mucosubstances. Metals that do not bind are expected to remain in the mouth for longer periods of time with more of an opportunity to permeate the linings and interact with cells of the mucosa. However, when one considers other bioevents, such as, the premature activation of enzymes by metals, which makes them useless for needed catalytic reactions, the interaction of metallic ions with the pellicle and plaque of teeth, and any possible absorption and toxication via the normal routes after swallowing, then strong binding promotes undesirable biochemistries. From the standpoint of the corrosion scientist, any time the products of corrosion are removed from the equilibrium confines of their source, energetics are created for the continuation of the corrosion. The binding of corroded metallic ions from dental alloys can, therefore, be regarded as being both an advantage and a liability. Efforts cer-

Binding of corroded ions: b1.J. Mueller

tainly need to be continued to establish an accountability of corroded metallic ions in the bioenvironment.

2

3

ACKNOWLEDGEMENTS

4

Partial financial support for this project was made possible through the Office of Sponsored Projects of the American Dental Association Health Foundation with a Biomedical Research Support Grant from the National Institute of Healths No. 54D-D2-13. Partial support was also obtained through a NIH grant with the Council on Dental materials No. DE45761.

5

6

7 8

REFERENCES 1

Dolby, A.E. (Ed.), Oral Mucosa in Health and Disease, Oxford, 1975, pp l-l 30

Blackwell,

9

Luckey, T.D., Venugopal, B. and Hutcheson, D., Heavy Metal Toxicity, Safety, and Hormology, Academic Press, New York, 1975, pp 4-l 2 Mueller, H.J., The binding of corroded metallic ions to salivay type proteins, Biomaterials 1983, 4. 66-72 Lowry, O.M., Rosebrough. N.J., Fan, A.L. and Randall, R.J., Protein measurement with the folin phenol reagent, J. &of. Chem. 3951, 193, 265 Ericson, T. Gel filtration of human saliva, Advances in Fioorine Research and Dental Caries Prevention 1966. 4. 169-l 80 169 Millin, D.J. and Smith, M.H.. Gel filtration and chromatography of human salivary proteins. Biochem. Biophys. Acta. 1962, 62, 45w55 Apostolopouios, A.X., On the tsolation of a human salivary giycoprotein, Archs. Oral Biol. 1967, 12, 789-797 Juriaanse, A.C. and Boou, M., Isolation and characterization of the main neutral protein from human submandibular saliva. Archs. Oral Biol. 1979. 24, 825-828 Ericson, T., Fractionation of glycoproteins on hydroxyapatite and on agarose gels, Ark&. Kemi. 1968, 29, 75-86

Biomateriais

f985,

Vof 6 May

f49