33
In Vitro Platelet Serotonin Secretion and Inhibition on Variously Treated Root Surfaces Arnold D.
Steinberg, * Jeannette Lipowski, * and Guy LeBreton*
Platelet
degranulation can result
in the release of
a
variety of factors which are
chemotactic, mitogenic, and angiogenic, making platelets extremely important in the regulation of the repair process. This study examines how various types of root surfaces affect platelet deposition and the release of serotonin from dense granules. In addition, experiments were performed to evaluate the effects of the cyclo-oxygenase inhibitor, indomethacin, on platelet deposition and dense granule release. Roots from freshly ex-
tracted teeth from sites with periodontal disease (PD) and from healthy sites were sectioned and had the following surface conditions: 1) periodontal ligament present; 2) PD; 3) PD, root planed; 4) PD, root planed and demineralized; and 5) condition 4 treated with collagenase. In addition, rabbit calcaneal tendon collagen was used. All samples were incubated with platelets labeled with both U1lndium and 14C serotonin, with and without the addition of indomethacin. It was observed that the greatest number of platelets deposited on the tendon collagen. Furthermore, serotonin release occurred on all samples except PD and indomethacin partially inhibited platelet deposition on all samples except tendon collagen. Finally, indomethacin inhibited serotonin release on all surfaces. These results suggest that attachment of platelets to the root surface is facilitated by metabolism through the cyclo-oxygenase pathway and that limited platelet deposition can occur in the absence of dense body release. J Periodontol 1992; 63:33-38.
Key Words: Indomethacin; blood platelets; serotonin; dense granule.
Root surface preparation is important in assisting the formation of new connective tissue attachment.14 This attachment is believed to be dependent upon the ability to form fibrin linkages to root surface collagen.5 We have suggested that root surface thrombogenicity is a necessity for fibrin linkage.6,7 It may, in fact, be the root surface thrombogenicity which determines the earliest events during healing at the soft tissue/root surface interface, initiating the cascade of events ultimately resulting in new connective tissue attachment. Hemostasis is the initial event in all wound healing. Our scanning electron microscopy (SEM) observations of clotting on various root surfaces suggested that the deposition of plasma proteins occurred first and was quickly followed by the adherence of aggregating platelets and the appearance of fibrin. For the most part, this fibrin most likely originated from fibrinogen present on the root surface from the initial plasma protein adsorption. Further interaction resulted in entrapment of numerous erythrocytes in a fibrin
'University of Illinois at Chicago, Department chemistry, Chicago, il. *Department of Pharmacology.
of Periodontics and Bio-
was rapidly followed by clot retraction and This entire process is very short-lived, the fibrinolysis.6'7 in seconds to minutes after injury. events occurring major The events associated with hemostasis may influence the second stage of healing, the inflammatory process.8 Platelets are the initial blood elements associated with hemostasis. Upon activation they synthesize arachidonic acid metabolites, primarily thromboxane A2 (TxA2), and release components from the various platelet granules.9 All of these products can affect not only hemostasis, but also the inflammatory process.10'11 For example, TxA2 is associated with bone résorption12 and platelet lysosomes contain hydrolytic enzymes similar to the lysosomal enzymes found in neutrophils.13'14 The release of fibrinogen, fibronectin, platelet factor 4, platelet-derived growth factor, and other products from the alpha granules may be important in wound healing.15 Dense granules may also be important in the progression of repair since they contain ADP and serotonin. ADP facilitates platelet recruitment. Serotonin, which is found in elevated concentrations in healing tissues, originates mainly from platelets.16 It is believed that serotonin promotes hemostasis by inducing vasoconstriction in small vessels11 and also has fibroblast stimulating properties.17
network which
J Periodontol
EFFECT OF ROOT SURFACES ON IN VITRO PLATELET SEROTONIN SECRETION
34
The current
investigation
is
a
continuation of
our
pre-
vious studies of platelet interaction with, and clot formation on, various tooth surfaces.6,7 It deals specifically with how various types of root surfaces affect platelet deposition and the release of serotonin by dense granules. Also investi-
gated was how indomethacin, an inhibitor of Prostaglandin metabolism and platelet activation, affects platelet deposition and dense granule release. Knowledge of the specific mechanisms involved in platelet attachment, accumulation, and granule release reactions on various root surfaces will provide an understanding of the initial events in periodontal wound healing and how various types of root treatments alter this process.
MATERIALS AND METHODS
Preparation of Root Samples and Tendon Collagen Freshly extracted human teeth had the crowns removed and the roots sectioned into approximately 2 mm thick, 5x5 squares, which included the root surfaces. Teeth from sites with periodontal disease (PD) that were to be used for mm
treatments
requiring root planing were planed prior to sec-
tioning. A total of 80 teeth with PD and 22 teeth without PD were used. A dissecting microscope was employed to confirm that sections of the root surface that were to be used were from portions of the root that were completely in either the PD root surface or in the periodontal ligament (PDL). Tooth selection, root preparation, and the sectioning procedure have been previously described.18 Tendon collagen was obtained from rabbit calcaneal tendons, lyophilized and stored at 80° C. -
Chemicals Used The source, concentration, and use of collagenase and citric acid (CA) are as previously described.6'18 Indium-Ill (mIn)Oxine* (aqueous) had a specific activity of 10 mCi/ug, and 14C serotonin (aqueous with 2% ethanol) had a specific activity of 54 mCi/mmole. Both reagents were obtained from Amersham.* Indomethacin was obtained from Sigma
Chemical.§
Surface Conditions to Be Tested All the sectioned samples were washed in 0.001 M phosphate buffer, pH 7.0, for 5 minutes and then placed in a 37° C oven for 20 minutes. These samples were prepared for each of the following treatments: 1. Intact PDL present (teeth from healthy sites): no treatment used. 2. PD present: no treatment used. 3. PD present: root planed. 4. PD present: root planed + 3-minutes CA incubation followed by a 3-minute sterile saline wash. *Amersham Co., Arlington Heights, IL. 5Sigma Chemical Co., St. Louis, MO.
January 1992
5. PD present: root planed + 3 minutes CA incubation followed by a 3-minute sterile saline wash and a 3-minute 0.001 M phosphate buffer (pH 7,0) wash. This was followed by a 5-minute, 37°C, incubation with collagenase. The samples were washed 3 times in 0.001 M phosphate buffer (pH 7) to stop the reaction. 6. Sections of the frozen, lyophilized rabbit tendon collagen were cut into pieces similar in size and shape to the root samples. These samples served as positive controls, since collagen is known to induce platelet activation and
release.19'20
All root samples and the collagen pieces were individually mounted in red dental wax exposing a 25 mm square area of the surface. All samples with similar surface treatments were incubated together in plastic vials. Prior to incubation with platelets, random samples of the various types of surfaces were examined using SEM to assure that proper surface consistency was maintained. During the course of these experiments, SEM was also performed on the various surface samples to observe the presence and morphological appearance of platelets (unpublished data). Platelet Rich Plasma (PRP) was obtained from healthy
volunteers who had not received any medication for 10
days.6
PRP
Labeling With inIn
and 14C
PRP, platelets were isolated by centrifugaUsing tion, resuspended in acid citrate dextrose-saline and incubated with mIn Oxine (50uCi) as previously described.6 The labeling efficiency of platelets with mIn was 74%. Labeled platelets were then isolated by centrifugation and the labeled pellet was resuspended in fresh PRP to a final volume of 80 ml. The average platelet count was now at 108 platelets/ml. This procedure resulted in mIn 3.6 labeled platelets with an average specific activity of 105 cpm/ 10s platelets of which 2% was free U1ln. This suspension (PRP) was then incubated with the second label, 14C serotonin (0.2 uCi/ml), for 30 minutes at 25° C. The labeling efficiency of platelets with 14C serotonin was 78%, yielding an average specific activity of 2 IO4 cpm/108 platelet. mIn and 14C serotonin were used as markers for the presence of platelets and release of dense granule constituents, respectively. This is a well-established methodology.6'2124 The red dental wax used to mount the samples was assessed to determine whether it stimulated platelet attachment and/or granule release. human
Platelet Incubation The samples for each of the conditions were incubated together in 5 ml of the U1ln and 14C -labeled PRP (3.6 x 108 platelets/ml) which was pre-incubated for 2 minutes, with and without 20 uM of indomethacin. An appropriate buffer blank was incorporated into the samples not receiving indomethacin to maintain equivalent concentrations. The samples were rinsed in saline to remove loosely adhering
Volume 63
Number 1
STEINBERG, LIPOWSKI, LeBRETON
platelets, removed from the wax housing, and the number of platelets deposited on the surfaces determined by counting for gamma activity of the '"In. Since the half-life of 11'In is 2.83 days, the samples were again counted for gamma activity 30 days later to assure that the niIn label had completely decayed. 14C associated with serotonin was then determined using a beta spectrometer. Parallel experiments, using free 11'In and free 14C serotonin in platelet free plasma, were performed to establish that the 11'In and 14C serotonin detected on the various surfaces was not the free isotope interacting with the sample surface. Data Analysis The 11'In served as a marker associated with the platelet membrane structure and represents the number of platelets which attach to each surface (Table 1, columns A and D). The 14C served as a marker for serotonin remaining in the dense granules and represents the number of platelets which retain their dense granules (Table 1, columns and F). If the number of platelets found on a specific surface were similar when determined by niIn and 14C, there was no 14C serotonin released. If the number of platelets determined by 14C content were found to be significantly less than the number of platelets found on the same surface as determined by inIn assay, 14C serotonin release had occurred. Student t tests were used to determine if these differences were
statistically significant.
RESULTS Column A in Table 1 demonstrates that platelets were deposited on all surfaces, with maximum platelet deposition Table 1. Platelet Attachment and Serotonin Release Indomethacin
Variously Treated
on
the tendon collagen surface. All surfaces, including tendon collagen, induced between 42% and 60% of serotonin release except for the PD surface (column C). Column D shows the effect of the presence of 20 uM of indomethacin on platelet deposition upon the various surfaces. It can be seen that indomethacin produced between 38% and 80% inhibition of platelet deposition on all root surfaces and had no such effect on tendon collagen (column E). The table also reveals that the addition of indomethacin significantly inhibited platelet serotonin release on all samples since the numbers of attached platelets with retained serotonin was in no case significantly different from the number of attached platelets (column G). Over a 1-hour time period no deposition of free 11'In or of free 14C was noted on the red dental wax control. Furthermore, the wax did not induce 11'In or 14C-serotonin release from the labeled platelets. Incubation of the various surfaces with free '"In resulted in approximately 0.01% of the indium becoming associated with the various sample surfaces. Since only 2% of the counts in the '"In labeled platelets were free, the amount of free '"In which bound to the various surfaces represents an insignificant portion (3 cpm) of the total counts observed in the presence of labeled platelets (2500 cpm). Similarly, incubation of the various surfaces with free 14C-serotonin resulted in approximately 0.009% of the free serotonin becoming associated with the surfaces. Again, this results in an insignificant portion (9 cpm) of the total counts observed in the presence of labeled platelets (800 cpm). Thus the possibility of erratic platelet counts from free "'In and free 14C-serotonin being released by the platelets during incuon
Root Surfaces and
Collagen With and Without
No Indomethacin
Surface Condition PDL Periodontal Disease (PD) PD +Planed
(PL)
Attached Platelets
the Addition of
Indomethacin Added
Attached Platelets With Retained Serotonin
% Serotonin Release
Attached Platelets
Inhibition
Attached Platelets With Retained Serotonin
(1-B/A)100
Mean SD
(1-D/A)100
Mean SD
A
vs
A
Mean SDf
N*
Mean
SD
0.65 0.42
0.42 0.27
32 28
0.38 0.37
0.23 0.14
20 20
42%* NS
0.33 0.26
0.27
0.27
32
0.12
0.11
20
56%*
vs
0.14 0.13
12 12
49%* 38%*
0.11
0.03
12
D
D
vs
(1-F/D)100
0.43 0.36
0.13 0.17
12 12
NS5 NS
59%*
0.15
0.03
15
NS
0.27
0.24
32
0.15
0.09
20
44%*
0.06
0.03
12
78%*
0.07
0.04
15
NS
0.35
0.18
36
0.14
0.06
24
60%*
0.07
0.03
12
80%*
0.10
0.02
15
NS
Collagenase Collagen
4.40
0.50
14
2.53
1.55
12
43%*
4.24
1.30
16
NS
3.61
1.66
16
NS
The samples for each of the conditions were incubated together in the double labeled PRP at 37°C, Uneven sample numbers between attached platelets and the serotonin values represent lost samples. All means expressed as platelets X 106/25mm2. 'Significantly different using student t test at (0.05). Standard deviation. +SD Number of samples. *N Not significant. SNS =
=
pH 7,
for 60 minutes.
F
% Serotonin Release
PD +PL + Citric Acid (CA) PD + PL + CA +
=
35
36
EFFECT OF ROOT SURFACES ON IN VITRO PLATELET SEROTONIN SECRETION
re-depositing on the various surfaces is an infactor. significant The use of SEM verified that platelets were attaching to the various surfaces and revealed various degrees of platelet activation (unpublished data). bation and
DISCUSSION We chose to investigate the involvement of the (recruitment) attachment process in platelet interaction with various root surfaces since this is an initiating event in wound healing, and may be linked to the maintenance of root surface attachment in vivo. When platelets are stimulated by tissue injury, contact with various chemical agents or surface interaction, an activation process results which can include the release of platelet granules. Four general responses to platelet interaction with a substrate have been identified, with the degree of the response varying: 1. The initial functional response is a shape change from discoid to spherical, with some pseudopod formation. Arachidonic acid metabolism does not appear to be involved with this phase. 2. Further platelet change is then followed by an increase in the adhesive properties of the platelet membranes resulting in platelet-platelet adhesion. Adhesion to the substrate may also occur and results in further pseudopod formation with spreading and flattening of the platelets over the exposed surface. During this process, platelet membrane phospholipase becomes activated resulting in stimulation of arachidonic acid metabolism and production of
TxA2.25
3. The TxA2 then platelet granules.9
causes
the secretion of the various
4. The contents of the dense granules, notably adenosine diphosphate (ADP), in turn recruit surrounding platelets to aggregate at the site of injury.25 The rationale for the use of the various root surface treatments in the present study was that they are associated with different periodontal healing. A root surface with PDL that has not been damaged will heal,26 whereas a root surface with PD seldom heals.27,28 A root planed PD surface will usually heal with a long junctional epithelium, whereas a PD root surface treated by root planing plus CA demineralization (at least in experimental animals) has success in new attachment.29 Collagenase treatment of the CA demineralized, root planed surface was used since in an earlier SEM study18 this treatment was efficient in elimination of the inter-collagenous material and appeared to result in the best exposure of the dentin fibrillar collagen. Since collagen is known to enhance platelet deposition and activation,19,20 our use of tendon collagen served as a positive control and a standard for other sample comparisons. In Table 1 we observed that between 42% to 60% of the platelet serotonin was released as a result of incubation with all surfaces except the PD surface. Finding no serotonin release on the PD surface was of interest since this is a
J Periodontol 1992
January
surface that does not facilitate healing.27,28 It is speculated that platelet release reaction may be involved in the progression of events leading to a healing of the periodontium and, for reasons presently unknown, PD engenders a root surface which does not facilitate a platelet activation response. Our results did not show that the collagenase-treated, CA-demineralized root surface enhanced either platelet deposition or serotonin release. This is in spite of the previous finding18 that SEM observations suggested better exposure of dentin fibrillar collagen is achieved under this condition. While both tendon collagen and PDL had similar percentages of serotonin released, the PDL was found to be associated with significantly fewer platelets. This result may be due to differences in the relative collagen exposure on each surface. Tendon collagen might be expected to provide a greater degree of platelet deposition because there is more collagen exposed per unit of surface than on roots. It should also be noted that platelet deposition on surfaces may involve two processes. One is the deposition or adhesion of platelets to the root surface itself; the other is platelet interaction with platelets already deposited on the surface. Since the current study does not differentiate between the two events, differences in total platelet deposition may be derived from either altered platelet-surface interaction or platelet-platelet interaction. The relationship between arachidonic acid metabolism and platelet deposition and dense granule secretion on the various surfaces was also assessed in this study. This was accomplished by using indomethacin, which inhibits the platelet cyclo-oxygenase pathway and thromboxane production.30 Table 1 demonstrates that the addition of 20 uM of indomethacin partially inhibits platelet deposition to the various root surfaces. This indicates that on these surfaces platelet deposition is, in part, controlled by the cyclo-oxygenase pathway. On the other hand platelet deposition on the tendon collagen surface appears to be independent of the cyclo-oxygenase pathway since deposition was not blocked by indomethacin treatment. These findings may be explained on the following basis: The requirement for the cyclo-oxygenase pathway in the platelet deposition process is dependent upon whether the surface is a strong platelet stimulator (collagen) or a weak platelet stimulator (ADP, epinephrine).31r34 Since TxA2 and associated secretions are amplification mechanisms to propagate an initial agonist or surface response, their involvement in platelet activation is more important when the agonist concentration is less than maximal.33 Thus at low collagen concentrations, as found on a root surface, the requirement for TxA2 and secretion would be more essential to the process of deposition than at high collagen concentrations such as associated with tendon collagen. The blocking of platelet secretion by indomethacin may be explained on the basis that cyclo-oxygenase inhibition is more effective in blocking platelet-platelet adhesion, or platelet aggregation,35,36 than in blocking
Volume 63 Number 1
platelet-collagen adhesion.37,38 Thus the effect of indomethacin on platelet-platelet interaction would be expected to be much greater than on platelet-collagen adhesion. Considering all of these findings together, the results indicate that while platelet deposition occurred on all of
these surfaces, the nature of the surface determines the number of platelets which attach and the degree of dense granule release. The results also suggest that attachment of platelets to the root surface is facilitated by metabolism through the cyclo-oxygenase pathway and that limited deposition can occur in the absence of dense body release. We hypothesize that only when platelet deposition is combined with activation and the resulting degranulation is this system poised for regeneration. It is, therefore, important that we understand the mechanisms which control platelet attachment, activation, and degranulation on various types of root surfaces. Such knowledge may enable us to control the initiating factor inducing the cascade of events which will finally result in the regeneration of a new connective tissue attachment on the root surface. Finding such a treatment could lead to more effective therapy for periodontal disease.
Acknowledgment
investigation was funded in part by The American Heart Association and The George Matula Fund and the Wach Fund, University of Illinois, College of Dentistry, Chicago, IL. We wish to thank Dr. June Steinberg for her helpful comments in the preparation of this manuscript. This
REFERENCES 1. Poison AM. The root surface and regeneration: Present therapeutic limitations and future biologic potentials. J Clin Periodontal 1986;
13:995-999. 2.
3.
4.
5.
6.
7. 8.
9.
10.
Lowenberg B, Thibault J, Lawrence C, Sodek J. The influence of chemically-induced modifications of root surfaces on cell migration, attachment and orientation. / Dent Res 1986; 65:1010-1014. Lai H, O'Leary TJ, Kafraway AH. The effect of different treatment
modalities on connective tissue attachment. J Periodontal 1986; 57:604612. Hancock EB. Regeneration procedures. In: Proceedings of the World Workshop in Clinical Periodontics. Chicago: The American Academy of Periodontology; 1989:Vl-l-Vl-20. Poison AM, Proye MP. Fibrin linkage: A precursor for new attachment. J Periodontol 1983; 54:141-147. Steinberg AD, LeBreton G, Willey R, Mukherjee S, Lipowski J. Extravascular clot formation and platelet activation on variously treated root surfaces. / Periodontol 1986; 57:516-522. Steinberg AD, Willey R. SEM observation of initial clot formation on treated root surfaces. J Periodontol 1988; 59:403^111. Silver MJ, Bills TK, Smith JB. Platelets and Prostaglandins: The key role of platelet phospholipase A2 activity. In: de Gaetano G, Garattini S, eds. Platelets: A Multidisciplinary Approach. New York: Raven Press: 1978:213-224. Hamber M, Svensson J, Samuelsson B. Thromboxanes: a new group of biologically active compounds derived from Prostaglandin endoperoxides. Proc Nati Acad Sci (USA) 1974; 72:3994-2998. Michaeli D, Hunt TK, Knighton DR. The role of platelets in wound healing: demonstration of angiogenic activity. In: Hunt TK, Heppen-
STEINBERG, LIPOWSKI, LeBRETON
37
stall RB, Pines E, Rovee D, eds. Soft and Hard Tissue Repair, Vol. 2. New York: Praeger Scientific, Surgical Science Series; 1984:380394. 11. Trowbridge HO and Emling RC. Inflammation. A Review of the Process. 3rd ed. Chicago: Quintessence Publishing Co; 1989:58-61. 12. Harvey W. Source of Prostaglandins and their influence on bone résorption and formation. In: Harvey W, Bennett A, eds. Prostaglandins in Bone Resorption. Boca Roton, FL: CRC Press; 1988:27-42. 13. Nachman R, Weksler B. The platelet as an inflammatory cell. In: Weissman G, ed. The Cell Biology of Inflammation. Amsterdam: Elsevier Biomedicai Press; 1980:145-161. 14. Marcus A, Zucker-Franklin D, Safier LB, Ullman L. Studies on human platelet granules and membranes. J Clin Invest. 1966; 45:1428. 15. White J. Ultrastructural modification in platelet membranes and cytoskeleton following activation. Blood Cell 1983; 9:237-261. 16. Castor CW. Autacoid regulation of wound healing. In: Glynn LE, ed. Tissue Repair and Regeneration. New York: Elsevier/North Holland Biomedicai Press; 1981:177-209. 17. Boucek RJ, Alvarez TR. 5-Hydroxytryptamine: A cyto specific growth stimulator of cultured fibroblasts. Science 1969; 167:898-899. 18. Willey RT, Steinberg AD. SEM studies of root dentin surfaces treated with citric acid, elastase, hyaluronidase, pronase and collagenase. / Periodontol 1984; 55:592-596. 19. Fromjovic MN, Milton J. Human platelet size, shape and related functions in health and disease. Physiol Rev 1982; 62:186-261. 20. Pachman MA, Mustard JF. Normal and abnormal platelet activity. In: Lasslo A, ed. Blood Platelet Function and Medicinal Chemistry. New York: Elsevier Biomedicai Press; 1984:61-128. 21. Ill CR, Engvall E, Ruoslahti E. Adhesion of platelets to laminin in the absence of activation. / Cell Biol 1984; 99:2140-2145. 22. Venton DL, Enke SE, LeBreton GC. Azaprostaoic acid derivative inhibitors of arachidonic acid induced platelet aggregation. / Med Chem 1979; 22:824-830. 23. Ito Y, Siado M, Imanishi Y. Adsorption of plasma proteins and adhesion of platelets onto novel polyetherureth aneureas-relationship between denaturation of adsorbed proteins and platelet adhesion. J Biomed Mater Res 1990;24:227-231. 24. Cholakis CH, Sefton MV. In vitro platelet interactions with a heparin polyvinyl alcohol hydrogel. / Biomed Mater Res 1989; 23:399^110. 25. Weiss HJ. Platelets: Pathophysiology and Antiplatelet Drug Therapy. New York; Alan R. Liss Inc.; 1978:1-28. 26. Nyman S, Houston F, Sarhed G, Lindhe J, Karring T. Healing following reimplantation of teeth subjected to root planing and citric acid treatment. / Clin Periodontol 1985; 12:294-305. 27. Lopez NJ, Belvederessi M, De la Sotta R. Inflammatory effects of periodontally diseased cementum studied by autogenous dental root implants in humans. J Periodontol 1980; 51:582-593. 28. Wirthlin MR. The current status of new attachment therapy. J Periodontol 1981; 52:529-544. 29. Cole RT, Crigger M, Egelberg J, and Selvig KA. Connective tissue regeneration to periodontally diseased teeth. / Periodont Res 1980; 15:1-7. 30. Lapetina G, Reep , Read NG, Moneada S. Adhesion of human platelet to collagen in the precsence of prostacylin, indomethacin and compound BW 755C. Thromb Res 1986; 37:325-335. 31. Santoro SA. Identification of a 160,000 dalton platelet membrane protein that mediates the initial divalent cation-dependent adhesion of platelets to collagen. Cell 1986; 46:913-920. 32. Seiss W, Cuatrecasas P, Lapetina EG. A role for cyclooxygenase products in the formation of phosphatidic acid in stimulated human platelets. Differential mechanism of action of thrombin and collagen. / Biol Chem 1983; 258:4683^1686. 33. Seiss W. Molecular mechanisms of platelet activation. Physiol Rev 1989; 69:58-178. 34. Lagas B. In-vitro platelet responses: Dense granule secretion. In:
J Periodontol
38
EFFECT OF ROOT SURFACES ON IN VITRO PLATELET SEROTONIN SECRETION
Holmsen H, ed. Platelet Responses and Metabolism, Vol I: Responses. Boca Raton, FL: CRC Press; 1986:115-143. 35. Zuker MB, Nachmias VT. Platelet activation. Arteriosclerosis 1985; 5:2-8. 36. Charo IF, Feinman RD, Detwiler TC. Interrelations of platelet aggregation and secretion. J Clin Invest 1977; 60:866-873. 37. Weiss H, Tschopp TB, Baumgartner HR. Impaired interaction (adhesion-aggregation) of platelets with the subendothelium in storage-pool disease and after aspirin ingestion. Engl J Med 1975; 293:619623.
38.
January 1992
Kinlugh-Rathbone , Cazenave J, Pachham M, Mustard JF. Effect of inhibitors of the arachidonate pathway on the release of granule contents from rabbit platelets adherent to collagen. Lab Invest 1980; 42:28-34.
Send reprint requests to: Dr. Arnold D. Steinberg, Department of Periodontology, University of Illinois, Dental School, 801 South Paulina, Chicago, IL 60612. Accepted for publication August 19, 1991.
In Vitro Platelet Serotonin Secretion and Inhibition on Variously Treated Root Surfaces Arnold D.
Steinberg, * Jeannette Lipowski, * and Guy LeBreton*
Platelet
degranulation can result
in the release of
a
variety of factors which are
chemotactic, mitogenic, and angiogenic, making platelets extremely important in the regulation of the repair process. This study examines how various types of root surfaces affect platelet deposition and the release of serotonin from dense granules. In addition, experiments were performed to evaluate the effects of the cyclo-oxygenase inhibitor, indomethacin, on platelet deposition and dense granule release. Roots from freshly ex-
tracted teeth from sites with periodontal disease (PD) and from healthy sites were sectioned and had the following surface conditions: 1) periodontal ligament present; 2) PD; 3) PD, root planed; 4) PD, root planed and demineralized; and 5) condition 4 treated with collagenase. In addition, rabbit calcaneal tendon collagen was used. All samples were incubated with platelets labeled with both U1lndium and 14C serotonin, with and without the addition of indomethacin. It was observed that the greatest number of platelets deposited on the tendon collagen. Furthermore, serotonin release occurred on all samples except PD and indomethacin partially inhibited platelet deposition on all samples except tendon collagen. Finally, indomethacin inhibited serotonin release on all surfaces. These results suggest that attachment of platelets to the root surface is facilitated by metabolism through the cyclo-oxygenase pathway and that limited platelet deposition can occur in the absence of dense body release. J Periodontol 1992; 63:33-38.
Key Words: Indomethacin; blood platelets; serotonin; dense granule.
Root surface preparation is important in assisting the formation of new connective tissue attachment.14 This attachment is believed to be dependent upon the ability to form fibrin linkages to root surface collagen.5 We have suggested that root surface thrombogenicity is a necessity for fibrin linkage.6,7 It may, in fact, be the root surface thrombogenicity which determines the earliest events during healing at the soft tissue/root surface interface, initiating the cascade of events ultimately resulting in new connective tissue attachment. Hemostasis is the initial event in all wound healing. Our scanning electron microscopy (SEM) observations of clotting on various root surfaces suggested that the deposition of plasma proteins occurred first and was quickly followed by the adherence of aggregating platelets and the appearance of fibrin. For the most part, this fibrin most likely originated from fibrinogen present on the root surface from the initial plasma protein adsorption. Further interaction resulted in entrapment of numerous erythrocytes in a fibrin
'University of Illinois at Chicago, Department chemistry, Chicago, il. *Department of Pharmacology.
of Periodontics and Bio-
was rapidly followed by clot retraction and This entire process is very short-lived, the fibrinolysis.6'7 in seconds to minutes after injury. events occurring major The events associated with hemostasis may influence the second stage of healing, the inflammatory process.8 Platelets are the initial blood elements associated with hemostasis. Upon activation they synthesize arachidonic acid metabolites, primarily thromboxane A2 (TxA2), and release components from the various platelet granules.9 All of these products can affect not only hemostasis, but also the inflammatory process.10'11 For example, TxA2 is associated with bone résorption12 and platelet lysosomes contain hydrolytic enzymes similar to the lysosomal enzymes found in neutrophils.13'14 The release of fibrinogen, fibronectin, platelet factor 4, platelet-derived growth factor, and other products from the alpha granules may be important in wound healing.15 Dense granules may also be important in the progression of repair since they contain ADP and serotonin. ADP facilitates platelet recruitment. Serotonin, which is found in elevated concentrations in healing tissues, originates mainly from platelets.16 It is believed that serotonin promotes hemostasis by inducing vasoconstriction in small vessels11 and also has fibroblast stimulating properties.17
network which
J Periodontol
EFFECT OF ROOT SURFACES ON IN VITRO PLATELET SEROTONIN SECRETION
34
The current
investigation
is
a
continuation of
our
pre-
vious studies of platelet interaction with, and clot formation on, various tooth surfaces.6,7 It deals specifically with how various types of root surfaces affect platelet deposition and the release of serotonin by dense granules. Also investi-
gated was how indomethacin, an inhibitor of Prostaglandin metabolism and platelet activation, affects platelet deposition and dense granule release. Knowledge of the specific mechanisms involved in platelet attachment, accumulation, and granule release reactions on various root surfaces will provide an understanding of the initial events in periodontal wound healing and how various types of root treatments alter this process.
MATERIALS AND METHODS
Preparation of Root Samples and Tendon Collagen Freshly extracted human teeth had the crowns removed and the roots sectioned into approximately 2 mm thick, 5x5 squares, which included the root surfaces. Teeth from sites with periodontal disease (PD) that were to be used for mm
treatments
requiring root planing were planed prior to sec-
tioning. A total of 80 teeth with PD and 22 teeth without PD were used. A dissecting microscope was employed to confirm that sections of the root surface that were to be used were from portions of the root that were completely in either the PD root surface or in the periodontal ligament (PDL). Tooth selection, root preparation, and the sectioning procedure have been previously described.18 Tendon collagen was obtained from rabbit calcaneal tendons, lyophilized and stored at 80° C. -
Chemicals Used The source, concentration, and use of collagenase and citric acid (CA) are as previously described.6'18 Indium-Ill (mIn)Oxine* (aqueous) had a specific activity of 10 mCi/ug, and 14C serotonin (aqueous with 2% ethanol) had a specific activity of 54 mCi/mmole. Both reagents were obtained from Amersham.* Indomethacin was obtained from Sigma
Chemical.§
Surface Conditions to Be Tested All the sectioned samples were washed in 0.001 M phosphate buffer, pH 7.0, for 5 minutes and then placed in a 37° C oven for 20 minutes. These samples were prepared for each of the following treatments: 1. Intact PDL present (teeth from healthy sites): no treatment used. 2. PD present: no treatment used. 3. PD present: root planed. 4. PD present: root planed + 3-minutes CA incubation followed by a 3-minute sterile saline wash. *Amersham Co., Arlington Heights, IL. 5Sigma Chemical Co., St. Louis, MO.
January 1992
5. PD present: root planed + 3 minutes CA incubation followed by a 3-minute sterile saline wash and a 3-minute 0.001 M phosphate buffer (pH 7,0) wash. This was followed by a 5-minute, 37°C, incubation with collagenase. The samples were washed 3 times in 0.001 M phosphate buffer (pH 7) to stop the reaction. 6. Sections of the frozen, lyophilized rabbit tendon collagen were cut into pieces similar in size and shape to the root samples. These samples served as positive controls, since collagen is known to induce platelet activation and
release.19'20
All root samples and the collagen pieces were individually mounted in red dental wax exposing a 25 mm square area of the surface. All samples with similar surface treatments were incubated together in plastic vials. Prior to incubation with platelets, random samples of the various types of surfaces were examined using SEM to assure that proper surface consistency was maintained. During the course of these experiments, SEM was also performed on the various surface samples to observe the presence and morphological appearance of platelets (unpublished data). Platelet Rich Plasma (PRP) was obtained from healthy
volunteers who had not received any medication for 10
days.6
PRP
Labeling With inIn
and 14C
PRP, platelets were isolated by centrifugaUsing tion, resuspended in acid citrate dextrose-saline and incubated with mIn Oxine (50uCi) as previously described.6 The labeling efficiency of platelets with mIn was 74%. Labeled platelets were then isolated by centrifugation and the labeled pellet was resuspended in fresh PRP to a final volume of 80 ml. The average platelet count was now at 108 platelets/ml. This procedure resulted in mIn 3.6 labeled platelets with an average specific activity of 105 cpm/ 10s platelets of which 2% was free U1ln. This suspension (PRP) was then incubated with the second label, 14C serotonin (0.2 uCi/ml), for 30 minutes at 25° C. The labeling efficiency of platelets with 14C serotonin was 78%, yielding an average specific activity of 2 IO4 cpm/108 platelet. mIn and 14C serotonin were used as markers for the presence of platelets and release of dense granule constituents, respectively. This is a well-established methodology.6'2124 The red dental wax used to mount the samples was assessed to determine whether it stimulated platelet attachment and/or granule release. human
Platelet Incubation The samples for each of the conditions were incubated together in 5 ml of the U1ln and 14C -labeled PRP (3.6 x 108 platelets/ml) which was pre-incubated for 2 minutes, with and without 20 uM of indomethacin. An appropriate buffer blank was incorporated into the samples not receiving indomethacin to maintain equivalent concentrations. The samples were rinsed in saline to remove loosely adhering
Volume 63
Number 1
STEINBERG, LIPOWSKI, LeBRETON
platelets, removed from the wax housing, and the number of platelets deposited on the surfaces determined by counting for gamma activity of the '"In. Since the half-life of 11'In is 2.83 days, the samples were again counted for gamma activity 30 days later to assure that the niIn label had completely decayed. 14C associated with serotonin was then determined using a beta spectrometer. Parallel experiments, using free 11'In and free 14C serotonin in platelet free plasma, were performed to establish that the 11'In and 14C serotonin detected on the various surfaces was not the free isotope interacting with the sample surface. Data Analysis The 11'In served as a marker associated with the platelet membrane structure and represents the number of platelets which attach to each surface (Table 1, columns A and D). The 14C served as a marker for serotonin remaining in the dense granules and represents the number of platelets which retain their dense granules (Table 1, columns and F). If the number of platelets found on a specific surface were similar when determined by niIn and 14C, there was no 14C serotonin released. If the number of platelets determined by 14C content were found to be significantly less than the number of platelets found on the same surface as determined by inIn assay, 14C serotonin release had occurred. Student t tests were used to determine if these differences were
statistically significant.
RESULTS Column A in Table 1 demonstrates that platelets were deposited on all surfaces, with maximum platelet deposition Table 1. Platelet Attachment and Serotonin Release Indomethacin
Variously Treated
on
the tendon collagen surface. All surfaces, including tendon collagen, induced between 42% and 60% of serotonin release except for the PD surface (column C). Column D shows the effect of the presence of 20 uM of indomethacin on platelet deposition upon the various surfaces. It can be seen that indomethacin produced between 38% and 80% inhibition of platelet deposition on all root surfaces and had no such effect on tendon collagen (column E). The table also reveals that the addition of indomethacin significantly inhibited platelet serotonin release on all samples since the numbers of attached platelets with retained serotonin was in no case significantly different from the number of attached platelets (column G). Over a 1-hour time period no deposition of free 11'In or of free 14C was noted on the red dental wax control. Furthermore, the wax did not induce 11'In or 14C-serotonin release from the labeled platelets. Incubation of the various surfaces with free '"In resulted in approximately 0.01% of the indium becoming associated with the various sample surfaces. Since only 2% of the counts in the '"In labeled platelets were free, the amount of free '"In which bound to the various surfaces represents an insignificant portion (3 cpm) of the total counts observed in the presence of labeled platelets (2500 cpm). Similarly, incubation of the various surfaces with free 14C-serotonin resulted in approximately 0.009% of the free serotonin becoming associated with the surfaces. Again, this results in an insignificant portion (9 cpm) of the total counts observed in the presence of labeled platelets (800 cpm). Thus the possibility of erratic platelet counts from free "'In and free 14C-serotonin being released by the platelets during incuon
Root Surfaces and
Collagen With and Without
No Indomethacin
Surface Condition PDL Periodontal Disease (PD) PD +Planed
(PL)
Attached Platelets
the Addition of
Indomethacin Added
Attached Platelets With Retained Serotonin
% Serotonin Release
Attached Platelets
Inhibition
Attached Platelets With Retained Serotonin
(1-B/A)100
Mean SD
(1-D/A)100
Mean SD
A
vs
A
Mean SDf
N*
Mean
SD
0.65 0.42
0.42 0.27
32 28
0.38 0.37
0.23 0.14
20 20
42%* NS
0.33 0.26
0.27
0.27
32
0.12
0.11
20
56%*
vs
0.14 0.13
12 12
49%* 38%*
0.11
0.03
12
D
D
vs
(1-F/D)100
0.43 0.36
0.13 0.17
12 12
NS5 NS
59%*
0.15
0.03
15
NS
0.27
0.24
32
0.15
0.09
20
44%*
0.06
0.03
12
78%*
0.07
0.04
15
NS
0.35
0.18
36
0.14
0.06
24
60%*
0.07
0.03
12
80%*
0.10
0.02
15
NS
Collagenase Collagen
4.40
0.50
14
2.53
1.55
12
43%*
4.24
1.30
16
NS
3.61
1.66
16
NS
The samples for each of the conditions were incubated together in the double labeled PRP at 37°C, Uneven sample numbers between attached platelets and the serotonin values represent lost samples. All means expressed as platelets X 106/25mm2. 'Significantly different using student t test at (0.05). Standard deviation. +SD Number of samples. *N Not significant. SNS =
=
pH 7,
for 60 minutes.
F
% Serotonin Release
PD +PL + Citric Acid (CA) PD + PL + CA +
=
35
36
EFFECT OF ROOT SURFACES ON IN VITRO PLATELET SEROTONIN SECRETION
re-depositing on the various surfaces is an infactor. significant The use of SEM verified that platelets were attaching to the various surfaces and revealed various degrees of platelet activation (unpublished data). bation and
DISCUSSION We chose to investigate the involvement of the (recruitment) attachment process in platelet interaction with various root surfaces since this is an initiating event in wound healing, and may be linked to the maintenance of root surface attachment in vivo. When platelets are stimulated by tissue injury, contact with various chemical agents or surface interaction, an activation process results which can include the release of platelet granules. Four general responses to platelet interaction with a substrate have been identified, with the degree of the response varying: 1. The initial functional response is a shape change from discoid to spherical, with some pseudopod formation. Arachidonic acid metabolism does not appear to be involved with this phase. 2. Further platelet change is then followed by an increase in the adhesive properties of the platelet membranes resulting in platelet-platelet adhesion. Adhesion to the substrate may also occur and results in further pseudopod formation with spreading and flattening of the platelets over the exposed surface. During this process, platelet membrane phospholipase becomes activated resulting in stimulation of arachidonic acid metabolism and production of
TxA2.25
3. The TxA2 then platelet granules.9
causes
the secretion of the various
4. The contents of the dense granules, notably adenosine diphosphate (ADP), in turn recruit surrounding platelets to aggregate at the site of injury.25 The rationale for the use of the various root surface treatments in the present study was that they are associated with different periodontal healing. A root surface with PDL that has not been damaged will heal,26 whereas a root surface with PD seldom heals.27,28 A root planed PD surface will usually heal with a long junctional epithelium, whereas a PD root surface treated by root planing plus CA demineralization (at least in experimental animals) has success in new attachment.29 Collagenase treatment of the CA demineralized, root planed surface was used since in an earlier SEM study18 this treatment was efficient in elimination of the inter-collagenous material and appeared to result in the best exposure of the dentin fibrillar collagen. Since collagen is known to enhance platelet deposition and activation,19,20 our use of tendon collagen served as a positive control and a standard for other sample comparisons. In Table 1 we observed that between 42% to 60% of the platelet serotonin was released as a result of incubation with all surfaces except the PD surface. Finding no serotonin release on the PD surface was of interest since this is a
J Periodontol 1992
January
surface that does not facilitate healing.27,28 It is speculated that platelet release reaction may be involved in the progression of events leading to a healing of the periodontium and, for reasons presently unknown, PD engenders a root surface which does not facilitate a platelet activation response. Our results did not show that the collagenase-treated, CA-demineralized root surface enhanced either platelet deposition or serotonin release. This is in spite of the previous finding18 that SEM observations suggested better exposure of dentin fibrillar collagen is achieved under this condition. While both tendon collagen and PDL had similar percentages of serotonin released, the PDL was found to be associated with significantly fewer platelets. This result may be due to differences in the relative collagen exposure on each surface. Tendon collagen might be expected to provide a greater degree of platelet deposition because there is more collagen exposed per unit of surface than on roots. It should also be noted that platelet deposition on surfaces may involve two processes. One is the deposition or adhesion of platelets to the root surface itself; the other is platelet interaction with platelets already deposited on the surface. Since the current study does not differentiate between the two events, differences in total platelet deposition may be derived from either altered platelet-surface interaction or platelet-platelet interaction. The relationship between arachidonic acid metabolism and platelet deposition and dense granule secretion on the various surfaces was also assessed in this study. This was accomplished by using indomethacin, which inhibits the platelet cyclo-oxygenase pathway and thromboxane production.30 Table 1 demonstrates that the addition of 20 uM of indomethacin partially inhibits platelet deposition to the various root surfaces. This indicates that on these surfaces platelet deposition is, in part, controlled by the cyclo-oxygenase pathway. On the other hand platelet deposition on the tendon collagen surface appears to be independent of the cyclo-oxygenase pathway since deposition was not blocked by indomethacin treatment. These findings may be explained on the following basis: The requirement for the cyclo-oxygenase pathway in the platelet deposition process is dependent upon whether the surface is a strong platelet stimulator (collagen) or a weak platelet stimulator (ADP, epinephrine).31r34 Since TxA2 and associated secretions are amplification mechanisms to propagate an initial agonist or surface response, their involvement in platelet activation is more important when the agonist concentration is less than maximal.33 Thus at low collagen concentrations, as found on a root surface, the requirement for TxA2 and secretion would be more essential to the process of deposition than at high collagen concentrations such as associated with tendon collagen. The blocking of platelet secretion by indomethacin may be explained on the basis that cyclo-oxygenase inhibition is more effective in blocking platelet-platelet adhesion, or platelet aggregation,35,36 than in blocking
Volume 63 Number 1
platelet-collagen adhesion.37,38 Thus the effect of indomethacin on platelet-platelet interaction would be expected to be much greater than on platelet-collagen adhesion. Considering all of these findings together, the results indicate that while platelet deposition occurred on all of
these surfaces, the nature of the surface determines the number of platelets which attach and the degree of dense granule release. The results also suggest that attachment of platelets to the root surface is facilitated by metabolism through the cyclo-oxygenase pathway and that limited deposition can occur in the absence of dense body release. We hypothesize that only when platelet deposition is combined with activation and the resulting degranulation is this system poised for regeneration. It is, therefore, important that we understand the mechanisms which control platelet attachment, activation, and degranulation on various types of root surfaces. Such knowledge may enable us to control the initiating factor inducing the cascade of events which will finally result in the regeneration of a new connective tissue attachment on the root surface. Finding such a treatment could lead to more effective therapy for periodontal disease.
Acknowledgment
investigation was funded in part by The American Heart Association and The George Matula Fund and the Wach Fund, University of Illinois, College of Dentistry, Chicago, IL. We wish to thank Dr. June Steinberg for her helpful comments in the preparation of this manuscript. This
REFERENCES 1. Poison AM. The root surface and regeneration: Present therapeutic limitations and future biologic potentials. J Clin Periodontal 1986;
13:995-999. 2.
3.
4.
5.
6.
7. 8.
9.
10.
Lowenberg B, Thibault J, Lawrence C, Sodek J. The influence of chemically-induced modifications of root surfaces on cell migration, attachment and orientation. / Dent Res 1986; 65:1010-1014. Lai H, O'Leary TJ, Kafraway AH. The effect of different treatment
modalities on connective tissue attachment. J Periodontal 1986; 57:604612. Hancock EB. Regeneration procedures. In: Proceedings of the World Workshop in Clinical Periodontics. Chicago: The American Academy of Periodontology; 1989:Vl-l-Vl-20. Poison AM, Proye MP. Fibrin linkage: A precursor for new attachment. J Periodontol 1983; 54:141-147. Steinberg AD, LeBreton G, Willey R, Mukherjee S, Lipowski J. Extravascular clot formation and platelet activation on variously treated root surfaces. / Periodontol 1986; 57:516-522. Steinberg AD, Willey R. SEM observation of initial clot formation on treated root surfaces. J Periodontol 1988; 59:403^111. Silver MJ, Bills TK, Smith JB. Platelets and Prostaglandins: The key role of platelet phospholipase A2 activity. In: de Gaetano G, Garattini S, eds. Platelets: A Multidisciplinary Approach. New York: Raven Press: 1978:213-224. Hamber M, Svensson J, Samuelsson B. Thromboxanes: a new group of biologically active compounds derived from Prostaglandin endoperoxides. Proc Nati Acad Sci (USA) 1974; 72:3994-2998. Michaeli D, Hunt TK, Knighton DR. The role of platelets in wound healing: demonstration of angiogenic activity. In: Hunt TK, Heppen-
STEINBERG, LIPOWSKI, LeBRETON
37
stall RB, Pines E, Rovee D, eds. Soft and Hard Tissue Repair, Vol. 2. New York: Praeger Scientific, Surgical Science Series; 1984:380394. 11. Trowbridge HO and Emling RC. Inflammation. A Review of the Process. 3rd ed. Chicago: Quintessence Publishing Co; 1989:58-61. 12. Harvey W. Source of Prostaglandins and their influence on bone résorption and formation. In: Harvey W, Bennett A, eds. Prostaglandins in Bone Resorption. Boca Roton, FL: CRC Press; 1988:27-42. 13. Nachman R, Weksler B. The platelet as an inflammatory cell. In: Weissman G, ed. The Cell Biology of Inflammation. Amsterdam: Elsevier Biomedicai Press; 1980:145-161. 14. Marcus A, Zucker-Franklin D, Safier LB, Ullman L. Studies on human platelet granules and membranes. J Clin Invest. 1966; 45:1428. 15. White J. Ultrastructural modification in platelet membranes and cytoskeleton following activation. Blood Cell 1983; 9:237-261. 16. Castor CW. Autacoid regulation of wound healing. In: Glynn LE, ed. Tissue Repair and Regeneration. New York: Elsevier/North Holland Biomedicai Press; 1981:177-209. 17. Boucek RJ, Alvarez TR. 5-Hydroxytryptamine: A cyto specific growth stimulator of cultured fibroblasts. Science 1969; 167:898-899. 18. Willey RT, Steinberg AD. SEM studies of root dentin surfaces treated with citric acid, elastase, hyaluronidase, pronase and collagenase. / Periodontol 1984; 55:592-596. 19. Fromjovic MN, Milton J. Human platelet size, shape and related functions in health and disease. Physiol Rev 1982; 62:186-261. 20. Pachman MA, Mustard JF. Normal and abnormal platelet activity. In: Lasslo A, ed. Blood Platelet Function and Medicinal Chemistry. New York: Elsevier Biomedicai Press; 1984:61-128. 21. Ill CR, Engvall E, Ruoslahti E. Adhesion of platelets to laminin in the absence of activation. / Cell Biol 1984; 99:2140-2145. 22. Venton DL, Enke SE, LeBreton GC. Azaprostaoic acid derivative inhibitors of arachidonic acid induced platelet aggregation. / Med Chem 1979; 22:824-830. 23. Ito Y, Siado M, Imanishi Y. Adsorption of plasma proteins and adhesion of platelets onto novel polyetherureth aneureas-relationship between denaturation of adsorbed proteins and platelet adhesion. J Biomed Mater Res 1990;24:227-231. 24. Cholakis CH, Sefton MV. In vitro platelet interactions with a heparin polyvinyl alcohol hydrogel. / Biomed Mater Res 1989; 23:399^110. 25. Weiss HJ. Platelets: Pathophysiology and Antiplatelet Drug Therapy. New York; Alan R. Liss Inc.; 1978:1-28. 26. Nyman S, Houston F, Sarhed G, Lindhe J, Karring T. Healing following reimplantation of teeth subjected to root planing and citric acid treatment. / Clin Periodontol 1985; 12:294-305. 27. Lopez NJ, Belvederessi M, De la Sotta R. Inflammatory effects of periodontally diseased cementum studied by autogenous dental root implants in humans. J Periodontol 1980; 51:582-593. 28. Wirthlin MR. The current status of new attachment therapy. J Periodontol 1981; 52:529-544. 29. Cole RT, Crigger M, Egelberg J, and Selvig KA. Connective tissue regeneration to periodontally diseased teeth. / Periodont Res 1980; 15:1-7. 30. Lapetina G, Reep , Read NG, Moneada S. Adhesion of human platelet to collagen in the precsence of prostacylin, indomethacin and compound BW 755C. Thromb Res 1986; 37:325-335. 31. Santoro SA. Identification of a 160,000 dalton platelet membrane protein that mediates the initial divalent cation-dependent adhesion of platelets to collagen. Cell 1986; 46:913-920. 32. Seiss W, Cuatrecasas P, Lapetina EG. A role for cyclooxygenase products in the formation of phosphatidic acid in stimulated human platelets. Differential mechanism of action of thrombin and collagen. / Biol Chem 1983; 258:4683^1686. 33. Seiss W. Molecular mechanisms of platelet activation. Physiol Rev 1989; 69:58-178. 34. Lagas B. In-vitro platelet responses: Dense granule secretion. In:
J Periodontol
38
EFFECT OF ROOT SURFACES ON IN VITRO PLATELET SEROTONIN SECRETION
Holmsen H, ed. Platelet Responses and Metabolism, Vol I: Responses. Boca Raton, FL: CRC Press; 1986:115-143. 35. Zuker MB, Nachmias VT. Platelet activation. Arteriosclerosis 1985; 5:2-8. 36. Charo IF, Feinman RD, Detwiler TC. Interrelations of platelet aggregation and secretion. J Clin Invest 1977; 60:866-873. 37. Weiss H, Tschopp TB, Baumgartner HR. Impaired interaction (adhesion-aggregation) of platelets with the subendothelium in storage-pool disease and after aspirin ingestion. Engl J Med 1975; 293:619623.
38.
January 1992
Kinlugh-Rathbone , Cazenave J, Pachham M, Mustard JF. Effect of inhibitors of the arachidonate pathway on the release of granule contents from rabbit platelets adherent to collagen. Lab Invest 1980; 42:28-34.
Send reprint requests to: Dr. Arnold D. Steinberg, Department of Periodontology, University of Illinois, Dental School, 801 South Paulina, Chicago, IL 60612. Accepted for publication August 19, 1991.
