15
Clinica Chimicu Acta, 96 (1979) 15-23 0 ElsevierlNorth-Holland Biomedical Press
CCA 1059
COMPARISON OF PLATELET COUNT AND PLATELET PROTEIN METHODS FOR DETERMINATION OF PLATELET MAO ACTIVITY
HERBERT JACKMAN, RAMESH ARORA and HERBERT Y. MELTZER * Department of Psychiatry, University of Chicago Pritzker School of Medicine, Illinois State Psychiatric Institute, 950 East 59th Street, Chicago, IL 60637 (U.S.A.) (Received December 26th, 1978)
Summary Platelet monoamine oxidase (MAO) activity in 10 normal volunteers was studied as a function of platelet protein or electronically-determined platelet counts. Comparisons of the two methods were made for samples assayed on the same day as well as one week later. The MAO activities resulting from both methods were significantly correlated and reproducible but the results of the platelet count method were, in most instances, slightly but significantly more reliable than the platelet protein method. The relevance of these results to the controversy concerning platelet MAO activity in schizophrenia is discussed.
Introduction Controversy as to whether platelet monoamine oxidase (MAO) activity is or is not significantly decreased in schizophrenics persists. Following the initial report of Murphy and Wyatt [l] of decreased platelet MAO activity in schizophrenics, there have been at least 37 reports [2-381 which concur with (N = 23) [2,4-6,8,10,11,15,16,18,20,24-27,30,32-381 or reject (N = 14) [3,7,9, 12-14,17,19,21-23,28,29,31] their original finding. Some of these reports include some of the same subjects. These studies have been reviewed in detail elsewhere [38-401. The reasons for the discrepancies in the investigation of the relationship of platelet MAO activity to schizophrenia has only been partially explored. Variations in patient population, including diagnostic criteria, differences in control groups, possible effects of stress, menstrual cycle and the many other factors which can affect platelet MAO activity [ 411 and differences in methods of analysis are among the likely factors that could cause these discrepant results. * To whom correspondence should be addressed.
16
This report will describe a controlled study of one aspect of the methodology used to measure platelet MAO activity. We and others have pointed out that the use of platelet protein to express the specific activity of platelet MAO could be an important source of the variance in results [19,20,24,40]. This method, while widely used in studies of MAO and other enzymes in platelets, is suspect because results can vary significantly with slight changes in the method of platelet pellet preparation due to contamination with proteins from other formed blood elements or trapped plasma [ 191. Alternatively, platelet MAO activity can be studied as a function of the number of platelets present in the reaction mixture [ 20,411. It has been suggested that this method would be a more reliable means of measuring platelet MAO activity [20,39,40]. However, no report exists to our knowledge of a direct comparison of the two methods. We report here an evaluation of the platelet count versus platelet protein as a means of determining the MAO activity of human platelets. Materials and Methods Blood collection and sample designation Venipuncture was performed on two occasions, 7 days apart, in the morning on ten male volunteers (21 to 35 yrs) who were in good health, had no family history of mental illness, and were drug free. Blood was withdrawn with a plastic syringe. The basic method utilized to determine MAO activity was that of Murphy et al. [ 391. We will describe our procedures in detail to permit replication of these studies, if desired. Eight and a half ml of blood was added to each of four polypropylene tubes (Falcon No. 2059) containing 1.5 ml acidcitratedextrose (ACD) anticoagulant. Each of the four tubes were thereafter handled as separate blood samples. Two tubes were used for the platelet count assay method and designated platelet count (PC)-1A and PC-1B. The remaining two tubes from the four obtained on day 1 were allocated for the protein determination assay method and designated platelet protein (PP)-1A and PP1B. Blood from the second drawing was handled in the same fashion and designated PCSA and PC-8B and PP-8A and PP-8B in analogy with the samples from the first blood sampling. Assay methods All sample preparation and MAO assays were performed on the same day the blood was drawn. Each tube containing 8.5 ml blood and 1.5 ml anticoagulant was thoroughly mixed and centrifuged at 300 X g for 10 min in a Sorval RC-5 centrifuge using a SS-34 rotor. The concentration of platelets in whole blood ranged from 2.2 to 3.4 X 10’ platelets/ml. Three ml of the platelet-rich plasma (PRP) from each of the four tubes from a given subject was transferred by plastic pipette to coded polypropylene tubes (Falcon No. 2036) taking great care not to disturb the settled red and white cells. The remaining blood was again thoroughly mixed and centrifuged at 600 X g for 10 min. One ml of the second PRP was added to the original PRP. Thus, there were two PRP samples of 4 ml each for the platelet count method and two for the platelet protein method from each day of the study. This procedure yielded approximately 57 + 9% of the original platelets present in whole blood. The concentration of
platelets in PRP generally ranged from 1.6 X lo8 platelets/ml to 2.8 X lo8 platelets/ml. Over 90% of the platelets were obtained in the first centrifugation.
Platelet count method The number of platelets for the platelet count method was determined in triplicate by electronic means (Coulter Thrombocounter, model TH) using a 10 ~1 sample of PRP diluted in 20 ml of Isoton@ II. This is the method of platelet counting which has been utilized by a majority of investigators reporting platelet MAO activity on a platelet count basis in schizophrenic patients [20, 24--27,30,34]. To determine the validity of the platelet count with our procedure, platelet counts were determined with the Thrombocounter and by skilled technicians in the clinical lab using a microscope and hemacytometer (Unopette@, B ec t on-Dickinson Co.). The PRP platelet counts of four samples determined by both methods were each within 4%. No WBCs or RBCs were noted in the visual fields we inspected. A 0.5-ml aliquot of the PRP was added to each of four polypropylene reaction tubes (Falcon No. 2063), each of which contained 0.4 ml preincubation solution. This preincubation solution consisted of 0.1 ml phosphate buffer (1.0 mol/l, pH 7.4) in each reaction tube; two of the four tubes had 0.1 ml of pargyline (2.4 mmol/l, in 1 mmol/l HCl) to inhibit platelet MAO; distilled water was used to bring the preincubation solution in each tube to 0.4 ml. After addition of the PRP to the preincubation solution, the tubes were mixed with a Vortex apparatus. The duplicate values for each sample were averaged and the result utilized for all subsequent analyses. The duplicates always agreed within 5%.
Protein determination method The 4 ml of PRP from PP-1A and 1B or PP-8A and 8B was centrifuged at 2400 X g for 10 min to produce a platelet pellet. The supematant was discarded and the platelet pellet was washed twice with l-ml portion of ice cold saline. Distilled water, 0.4 ml, at ice bath temperature, was added to the pellet which was then sonicated for 10 set (setting 4, model S75-Heat Systems-Ultrasonics, Inc.) over ice. In an initial experiment, platelets were electronically counted in samples before and after sonication. The platelet count decreased by 98.2% ? 1.9% (N = 11) reflecting an almost complete disruption of the platelets by sonication. To assay MAO activity by the platelet protein method, four tubes were prepared for each sonicate in the manner described above with the exception that water was added to all tubes to a volume of 0.85 ml. To these reaction tubes, a 50-~1 aliquot of the platelet sonicate was added, following which the tubes were mixed with a Vortex@ apparatus. Both sets of reaction tubes were then preincubated at the same time in a shaking water bath at 37°C for 10 min in order for pargyline to inhibit platelet MAO. All the reaction tubes were then placed in an ice water bath for five min to temporarily halt the remaining MAO activity. One hundred ~1 of [14C]benzylamine (BA) (4.0 mmol/l, 600 &i/mmol, ICN Pharmaceuticals) was then quickly added to all the reaction tubes while in the ice bath, after which mixing was done with a Vortex@ apparatus. All the reaction tubes were then returned to the shaking water bath (37°C). After a 30-min incubation, all the tubes were again returned to the ice bath for 5 min so as to halt MAO activity, 0.3 ml of
18
3.0 mol/l HCI was then added and the tubes mixed with a Vortex@ apparatus to irreversibly inhibit the platelet MAO reaction. Three ml of heptane were added to all the reaction tubes which were then capped and mechanically shaken for 10 min to extract labelled reaction product. After shaking, the reaction tubes were centrifuged to break the emulsion and then placed in an acetone-dry ice bath to freeze the aqueous layer, allowing the heptane-product containing layer to be poured into counting vials. The amount of 14C-activity, representing the reaction products present in each heptane extraction, was determined with a liquid scintillation counter as disintegrations per min (dpm) per ml of reaction mixture. Recovery of the labelled reaction products was estimated by re-extracting samples with a second 3.0-ml portion of heptane. The total protein present in 50 ~1 of PS was determined by the method of Lowry et al. {421. Disintegrations per min due to platelet MAO activity were calculated as the difference between the average of tubes with and without pargyline [39]. Platelet MAO activity is reported as nmol of product per h per lo8 platelets or as nmol of product per h per mg protein for the platelet count and platelet protein methods respectively. Measures of reliability The reliabilities of either the platelet count assay method or the protein determination assay method were determined by calculating the coefficients of variation (C.V.) and intraclass correlations (I.C.C.) [43,44]. Re~~a~~lityof assay measures from same day duplicates Coefficients of variation and the I.C.C. were determined for the various assay measures obtained for the samples PC-1A and PC-1B and PCSA and PC8B from each subject. The measures which were evaluated were dpmiml reac tion volume, platelets/ml PRP and platelet MAO activity. A comparable series of C.V. and an I.C.C. were calculated using the same measures obtained with each subject’s PP-1A and PP-1B and PP-8A and PP-8B samples with the exception that protein/50 ~1 sonicate was used rather than platelets/ml PRP. The C.V. and I.C.C. formulated from these samples constitute the “same day pairs for day one” and are presented in Tables I-IV. Differences between the C.V. for the two methods were evaluated by means of a paired t-test. The C.V. and ICC, were also formulated for the same assay measures generated from the PC-8A and PC-8B, and from PP-8A and PP-8B for each subject and constitute the “same day pairs for day eight”. These results are also presented in Tables I-IV. Reliability of assay measures for assays done at seuen day i~teruals To compare the reliability of the measures obtained in blood samples from the same subject obtained on day one and on day eight, C.V. and I.C.C. were formulated from the means of the assay measures obtained for PC-1A and PC1B and for the means of the measures from PC-8A and PC-8B. The C.V. and I.C.C. were also formulated using the means of the assay measures described above from PP-1A and PP-lB, PP-8A and PP-8B. These C.V. and I.C.C. constitute the values referred to as “samples at seven day intervals” in Tables I-IV.
19
Results The mean platelet MAO activity using the platelet count assay was 6.89 t 1.66 units (mean kS.D.) for the 10 samples obtained on day one and 7.07 + 1.83 units for samples from the same 10 subjects obtained seven days later. The mean platelet MAO activity using the protein determination assay was 63.5 k 21.0 units for lo-samples from day one and 59.8 f 15.9 units for samples from the same 10 subjects seven days later, on day eight. These levels are within the normal limits for both assays in our laboratory. The Pearson correlation coefficients between the MAO activity determined by the platelet protein and platelet count methods using means of the pairs for each method from each subject for day 1 and day 8 were 0.775 and 0.762, respectively (p < 0.01). The mean C.V. for the same day pairs using the platelet count method from day one and day eight are 2.97% k 2.47% and 2.91 k 1.74%, respectively. Both values are significantly smaller 03 < 0.05) than the corresponding mean C.V. for day one and day eight, 5.62% + 4.35% and 5.43% + 3.89%, respectively, which were obtained with the protein determination method (Table I). The mean C.V. of the two methods for the samples obtained at an interval of seven days were not significantly different, in part because of the large variance in the platelet protein samples. The mean C.V. for the platelet count method was less than half that of the protein method. There was as much as a 30% discrepancy in the samples obtained at a ‘I-day interval with the platelet protein method whereas no subject had MAO levels which differed by more than 11% with the platelet count method. The I.C.C. for all day one or all day eight pairs was highly significant and virtually identical for both methods (0.979 and 0.981 for the platelet count method and 0.970 and 0.952 for the platelet protein method) (Table II). The I.C.C. based on the mean of split duplicates from day one and day eight of the study was considerably better for the platelet count method, 0.941, than for samples assayed using the protein determination method, 0.852 (Table II). However, both were significant at the 0.01 level. To further identify the possible contribution of variance in both assays, we
TABLE I PLATELET MAO ACTIVITY DETERMINATIONS (C.V.) IN TEN MALE SUBJECTS
AND
THEIR
COEFFICIENTS
Platelet assay
Same day pairs Mean day 1 Mean day 8
OF VARIATION
Protein *ssay
Activity *
C.V. (96)
6.89 + 1.66 7.07 t 1.83
2.97 f 2.47 2.91 * 1.74
6.98 r 1.69
4.25 f 3.63
9 5
Activity **
C.V. (%)
63.5 + 21.0 59.8 f 15.9
5.62 + 4.35 5.43 r 3.89
61.6 * 18.2
8.93
Samples at a ‘I-day interval Using mean values for day 1 and day 8
* nmol product/lo8 platelets (mean f S.D.). ** nmol product/h/mg protein (mean f S.D.). 5 Coefficients of variation are significantly different, P < 0.05 by paired t-test.
? 9.76
5 8
20 TABLE
II
INTRACLASS
CORRELATIONS
Same day pairs Day 1 Day 8 Samples at a 7-day interval * nmol product/h/lo8 * * nmol product/h/mg ***p < 0.01.
FOR PLATELET
MAO ACTIVITY
IN TEN MALE
Platelet assay *
Protein assay * *
0.979 0.981 0.941
0.970 0.952 0.852
*** *** ***
SUBJECTS
*** *** ***
platelets. protein.
examined the reliability of the various component measures of the platelet MAO activity. In Tables III and IV, the C.V. and I.C.C. of the dpm/ml of reaction volume, the platelets/ml of PRP and the protein/50 ~1 of sonicate are compared. The C.V. for dpm/ml of reaction volume were not significantly different for day one pairs, for day eight pairs or for samples obtained at a seven day interval for the two assay methods (Table III). The C.V. for platelets/ml of PRP and protein/50 ~1 of sonicate of day one and day eight pairs (Table III) are significantly different: the platelet counts from both days are significantly more reliable than the corresponding protein determinations (p < 0.025). There was no significant difference between C.V. of platelet counts or protein determinations in samples obtained at the seven day interval (Table III). The data of Table IV indicates that the I.C.C. for dpm/ml of reaction volume and for platelets/ml were significant at p < 0.01 while the I.C.C. for protein/50 1_r1 of sonicate determinations were significant at p < 0.05. We also compared the dpm/h/ml of reaction volume for the two methods.
TABLE
III
COEFFICIENTS
OF VARIATION
FOR COMPONENT
MEASURES
OF PLATELET
C.V. (%) Platelet assay
Protein assay
dpm/ml of reaction volume Same day pairs Mean day 1 Mean day 8 Samples at a 7-day interval
5.68 ? 4.89 4.06 k 3.39 7.35 k 7.60
6.16 6.61 7.98
Platelets/ml of PRP Same day pairs Mean day 1 Mean day 8 Samples at a ‘T-day interval
2.01 2.53 9.10
-
Protein/50 ~1 of sonicate Same day pairs Mean day 1 Mean day 8 Samples at a ‘T-day interval
-
* P < 0.025
when compared
? 2.12 * + 1.14 * ? 6.66
to same day split duplicates
8.46 8.06 9.96
2 4.23 f 5.91 it 6.37
+ 6.67 + 3.26 + 5.63
of protein150
~1 of sonicate.
MAO
ACTIVITY
21 TABLE
IV
INTRACLASS
CORRELATIONS
FOR
COMPONENT
MEASURES
OF PLATELET
MAO ACTIVITY
I.C.C. Platelet
assay
Protein
dpm/ml of reaction volume Same day pairs Mean day 1 Mean day 8 Samples at a 7-day interval
0.967 0.974 0.853
* * *
0.964 0.942 0.939
Platelets/ml of PRP Same day pairs Mean day 1 Mean day 6 Samples at a 7-day interval
0.986 0.989 0.776
* * *
-
Protein/50 ~1 of sonicate Same day pairs Mean day 1 Mean day 8 Samples at a ‘I-day interval
-
0.590 0.835 0.596
assay
* * *
** * **
* P < 0.01. * * P < 0.05.
This comparison is meaningful because the reaction volume and the reagent concentrations are the same in both assay methods while the quantity of platelet MAO protein in 50 ~1 of sonicate used in the platelet protein assay should be the same as the quantity of platelet MAO protein contained in 0.5 ml PRP. We found that the mean platelet MAO activity in the sonicate was 28.5% 2 12.7% less than that in the PRP used for the platelet count assay in samples from day one and 32.9% ?r 4.7% less in determinations from day eight samples. Combining the results of both days, the mean dpm/h/ml of reaction volume was 30.8% + 7.5% less for the protein assay than for the platelet count assay (p < 0.001, paired t-test). Discussion The results reported here indicate that the reliabilities of the platelet MAO activity and the component measures of that activity in same day paired samples for the platelet count assay method tend to be greater than that of the protein assay method. While the absolute differences in assay reliabilities are not great, every type of comparison we employed reflected a greater reliability for the platelet count method than for the protein method. Our results should be considered a minimum difference between the two methods because of the great care which we applied to all aspects of these determinations, including assaying each sample in duplicate. A probable cause for the lesser reliability of the protein assay versus the platelet count assay is a variable contribution of protein from red and white blood cells plus other noncellular proteins and aromatic amino acids which are measured along with platelet protein [ 19,30,39,40]. This conclusion is partially
22
supported by our data which shows no significant differences in reliability between the dpm/ml reaction volume generated by either assay method while the variability of the total protein determination was substantially greater than the variability of the platelet counts for same day duplicates (Tables III and IV). Since we found no WBCs or RBCs in the PRP which we examined in the microscope, we can assume that non-cellular protein is the source of the contaminating protein. Our results also indicate that there are significant differences in the MAO activity which can be obtained from the same PRP by the two methods, The loss of approximately 30% of MAO activity as a result of the sonication which is employed in the protein determination method suggests that the platelet count method could provide a more valid estimate of platelet MAO activity than the protein determination method. The majority of studies of platelet MAO activity in schizophrenics have employed the somewhat less reliable platelet protein method [l-10,14-24,28,29, 31--33,35,36]. Of these 28 studies, 16 reported decreased platelet MAO in schizophrenics. Seven of nine studies with the platelet count method reported decreased platelet MAO activity (Chi square = 0.512). Thus, there is little indication that the studies which utilize the platelet count method demonstrate a lower platelet MAO activity in schizophrenics more frequently than studies which utilize the platelet protein method. If done with care, the platelet protein method can achieve acceptable reliability but is is very easy for large errors to occur because of the effects of sonication and the possible trapping of nonplatelet protein in the pellet [19]. In our hands, the protein determination showed an average 8.3% i 3.8% variation in same day paired samples. The average C.V. for electronic platelet counting was 2.2% ?: 1.2% for same day pairs. Therefore, we believe the platelet count method for measuring platelet MAO activity, with electronic platelet counting, is the preferred method for determining platelet MAO activity. Our current investigation of platelet MAO activity in schizophrenics utilizing the platelet count method suggests that there is a subgroup of schizophrenics, characterized by hallucinations, who have signific~tly lower platelet MAO activity than non-hallucinating schizophrenics and normal controls (Meltzer, H.Y., Arora, R. and Jackman, H., in preparation). This is in agreement with previous reports of Schildkraut et al. [ 181 and our laboratory [24] and suggests that patient population may be a significant factor explaining differences between studies. Acknowledgements This study was supported in part by USPHS 25 116 and MHCRC 30 938. H.Y.M. is recipient of TJSPHS RCSA MH 47 808. References 1 Murphy. D. and Wyatt, R. (1972) Nature 238, 225-226 2 Wyatt, R., Murphy, D., Belmaker. R., Cohen. S., Donnelly, 916filS
C. and Pollin, W. (1973)
Science 179.
23 3 Friedman, E., Shopsin, B., Sathononthan, G. and Gershon. S. (1974) Am. J. Psychiatr. 131, 13921394 4 Meltsex, H.Y. and Stahl, S. (1974) Res. Common. Chem. PathoI. Pharmacd. 7.419431 5 Nies. A., Robinson, D., Harris, L. and Lamborn, K. (1974) in Advances in Biochemical Psychopharmacology (Usdin, E.. ed.), pp. 59-70, Raven Press, New York 6 Owen, F., Ridges, P. and Cookson, I. (1973) Acta Genet. Med. GemeIIol. (Roma) 23.371-376 ‘7 Carpenter. W.. Murphy, D. and Wyatt, R. (1975) Am. J. Psychiatr. 132.438441 8 Cookson, I., Owen, F. and Ridges, A. (1975) Psychol. Med. 5. 314-316 9 Shaskan, E. and Becker, R. (1975) Nature 253.659-660 10 Takahashi, S., Yamane. H. and Toni, N. (19753 FolIa Psychiatr. Neurol. Jap. 29.207-214 11 ZeIIer, E., Bashes, B., Davis, J. and Thorner, M. (1975) Lancet i, 1385 12 Belmaker. R., Ebstein, R., Rimon, R., Wyatt. R. and Murphy, D. (1976) Acta Psychiatr. Stand. 54. 67-72 13 Belmaker, R., Ebbesen, K.. Ehstein, R. and Rimon, R. (1976) Br. J. Psycbiatr. 129, 227-232 14 Brockington, I., Crow, T., Johnstone, E. and Owen, F. (1976) in Monoamine Oxidase and its inhibitors (Ciba Inc., ed.), pp. 353-362, Elsevier, New York 15 Domino, E. and Khanna, S. (1976) Am. J. Psychiatr. 133, 323-326 16 Murphy. D., Donnelly, C.. Miller, L. and Wyatt, R. (1976) Arch. Gen. Psychiatr. 33.1377-1381 17 Owen. F., Bourne, R., Crows, T., Johnstone, E., Bailey, A. and Her&ion, H. (1976) Arch. Gen. PSYchiatr. 33.1370-1313 18 SchiIdkraut, I., Herzog, J., Orsuiak, P., Edelman. S.. Stein, H. and Frazier, S. (1976) Am. J. Psychiatr. 133,438--440 19 White, H., McLeod. M. and Davidson. J. (1976) Am. J. Psychiatr. 133,1191-1193 20 Wyatt, R. and Murphy. D. (1976) Schiz. Bull. 2.77-89 21 Becker, R. and Shaskan, E. (1977) Am. J. Psychiatr. 134,512--517 22 Belmaker, R., Galon, A., Perez, L. and Ebstein, R. (1977) Br. 3. Psychiatr. 131. 551-552 23 Joseph, M., Owen, F.. Baker, H. and Boume, R. (1977) PsychoI. Med. 7.159-162 24 Meltzer, H.. Pscheidt, G., Goode, D.. PiyakaImaIa, S. and Dichtel, M. (1377) presented at the American Psychiatric Association Meeting, Toronto. Canada, May, 1977 25 Murphy, D.. Bebnaker, R., Carpenter, W. and Wyatt, R. (1977) Br. J. Psychiatr. 130.151-158 26 Sullivan, J., StanfieId,C. and Da&is, C. (1977) Am. J. Psychiatr. 134.1098-1103 27 Berger, P.. Ginsberg, R., Barchas, J., Murphy, D. and Wyatt, R. (1978) Am. J. Psychiatr. 135.9549 28 Friedhoff, A., Miller, J. and Weisenfreund, J. (1978) Am. J. Psyehiatr. 135.952-955 29 Groshong. R., Baldesstini, R., Gibson, D.. Lipinski, J., AxeIrod, D. and Pope, H. (19781 Arch. Gen. Psychiatr. 135,1198-1205 30 Potkin, S,. Cannon, H., Murphy, D. and Wyatt, R. (19781 New Eng. J. Med. 298.61-66 31 Schildkraut, J.. Orsulak. P.. Schatzberg, A., Cole, J., Gudeman. J. and Rhode, W. (1978) Am. J. PSYchiatr. 135,110-112 32 SuIIivan, J., Cavenar, J., Stanfield, C. and Hammett, E. (1978) Am. J. Psychiatr. 135. 597-698 33 Valkenberg, C. and Crowe, R. (1978) New Eng. J. Med. 298, 1150 34 Van Kammen, I).. Marder. S., Murphy, D. and Burmey, W. (19781 Am. 3. Psychiatr. 35.600-605 35 Berrettini, W., Prozialeck, W. and Vogel, W. (1978) Arch. Gen. Psychiatr. 35, 600-605 36 Berrettini, W. and Vogel, W. (1978) Am. J. Psychiatr. 135.605-607 37 SuIIivan. J., MaItbie. A., Cavenar, J.. Hammett, E. and Lister, P. (1978) New Eng. J. Med. 298,115l 38 Wyatt, R., PotkIn. S.. GiIbin, J. and Murphy, D. (1978) in Psychopharmacology: A Generation of Progress (Lipton, M., et al., eds.), pp. 1093-1095, Raven Press, New York 39 Murphy, D., Wright, C., Buchsbaum, M., Nichols. A., Costa, J. and Wyatt, R. (1976) Biochem. Med. 16.254-265 40 Meltzer. H.Y. (1978) in The Schizophrenia Syndrome (BeIIak, L., ed.), Grune and Stratton, New York, in press 41 Youdim, M.B.H. and Holzbauer. M. (1976) J. Neurrd Trans. 38.193-229 42 Lowry, O.H., Rosebrough, N.J., Farr. A.L. and RandaIl. R.J. (1951) J. Biol. Chem. 193,265-275 43 Snedecor, G. and Cochran, W. (1967) Statistical Methods, 6th edn., Iowa State University Press, Ames, Iowa 44 Bartko, J. and Carpenter, W. (1076) J. Nerv. Ment. Dis. 163.307-317
Clinica Chimicu Acta, 96 (1979) 15-23 0 ElsevierlNorth-Holland Biomedical Press
CCA 1059
COMPARISON OF PLATELET COUNT AND PLATELET PROTEIN METHODS FOR DETERMINATION OF PLATELET MAO ACTIVITY
HERBERT JACKMAN, RAMESH ARORA and HERBERT Y. MELTZER * Department of Psychiatry, University of Chicago Pritzker School of Medicine, Illinois State Psychiatric Institute, 950 East 59th Street, Chicago, IL 60637 (U.S.A.) (Received December 26th, 1978)
Summary Platelet monoamine oxidase (MAO) activity in 10 normal volunteers was studied as a function of platelet protein or electronically-determined platelet counts. Comparisons of the two methods were made for samples assayed on the same day as well as one week later. The MAO activities resulting from both methods were significantly correlated and reproducible but the results of the platelet count method were, in most instances, slightly but significantly more reliable than the platelet protein method. The relevance of these results to the controversy concerning platelet MAO activity in schizophrenia is discussed.
Introduction Controversy as to whether platelet monoamine oxidase (MAO) activity is or is not significantly decreased in schizophrenics persists. Following the initial report of Murphy and Wyatt [l] of decreased platelet MAO activity in schizophrenics, there have been at least 37 reports [2-381 which concur with (N = 23) [2,4-6,8,10,11,15,16,18,20,24-27,30,32-381 or reject (N = 14) [3,7,9, 12-14,17,19,21-23,28,29,31] their original finding. Some of these reports include some of the same subjects. These studies have been reviewed in detail elsewhere [38-401. The reasons for the discrepancies in the investigation of the relationship of platelet MAO activity to schizophrenia has only been partially explored. Variations in patient population, including diagnostic criteria, differences in control groups, possible effects of stress, menstrual cycle and the many other factors which can affect platelet MAO activity [ 411 and differences in methods of analysis are among the likely factors that could cause these discrepant results. * To whom correspondence should be addressed.
16
This report will describe a controlled study of one aspect of the methodology used to measure platelet MAO activity. We and others have pointed out that the use of platelet protein to express the specific activity of platelet MAO could be an important source of the variance in results [19,20,24,40]. This method, while widely used in studies of MAO and other enzymes in platelets, is suspect because results can vary significantly with slight changes in the method of platelet pellet preparation due to contamination with proteins from other formed blood elements or trapped plasma [ 191. Alternatively, platelet MAO activity can be studied as a function of the number of platelets present in the reaction mixture [ 20,411. It has been suggested that this method would be a more reliable means of measuring platelet MAO activity [20,39,40]. However, no report exists to our knowledge of a direct comparison of the two methods. We report here an evaluation of the platelet count versus platelet protein as a means of determining the MAO activity of human platelets. Materials and Methods Blood collection and sample designation Venipuncture was performed on two occasions, 7 days apart, in the morning on ten male volunteers (21 to 35 yrs) who were in good health, had no family history of mental illness, and were drug free. Blood was withdrawn with a plastic syringe. The basic method utilized to determine MAO activity was that of Murphy et al. [ 391. We will describe our procedures in detail to permit replication of these studies, if desired. Eight and a half ml of blood was added to each of four polypropylene tubes (Falcon No. 2059) containing 1.5 ml acidcitratedextrose (ACD) anticoagulant. Each of the four tubes were thereafter handled as separate blood samples. Two tubes were used for the platelet count assay method and designated platelet count (PC)-1A and PC-1B. The remaining two tubes from the four obtained on day 1 were allocated for the protein determination assay method and designated platelet protein (PP)-1A and PP1B. Blood from the second drawing was handled in the same fashion and designated PCSA and PC-8B and PP-8A and PP-8B in analogy with the samples from the first blood sampling. Assay methods All sample preparation and MAO assays were performed on the same day the blood was drawn. Each tube containing 8.5 ml blood and 1.5 ml anticoagulant was thoroughly mixed and centrifuged at 300 X g for 10 min in a Sorval RC-5 centrifuge using a SS-34 rotor. The concentration of platelets in whole blood ranged from 2.2 to 3.4 X 10’ platelets/ml. Three ml of the platelet-rich plasma (PRP) from each of the four tubes from a given subject was transferred by plastic pipette to coded polypropylene tubes (Falcon No. 2036) taking great care not to disturb the settled red and white cells. The remaining blood was again thoroughly mixed and centrifuged at 600 X g for 10 min. One ml of the second PRP was added to the original PRP. Thus, there were two PRP samples of 4 ml each for the platelet count method and two for the platelet protein method from each day of the study. This procedure yielded approximately 57 + 9% of the original platelets present in whole blood. The concentration of
platelets in PRP generally ranged from 1.6 X lo8 platelets/ml to 2.8 X lo8 platelets/ml. Over 90% of the platelets were obtained in the first centrifugation.
Platelet count method The number of platelets for the platelet count method was determined in triplicate by electronic means (Coulter Thrombocounter, model TH) using a 10 ~1 sample of PRP diluted in 20 ml of Isoton@ II. This is the method of platelet counting which has been utilized by a majority of investigators reporting platelet MAO activity on a platelet count basis in schizophrenic patients [20, 24--27,30,34]. To determine the validity of the platelet count with our procedure, platelet counts were determined with the Thrombocounter and by skilled technicians in the clinical lab using a microscope and hemacytometer (Unopette@, B ec t on-Dickinson Co.). The PRP platelet counts of four samples determined by both methods were each within 4%. No WBCs or RBCs were noted in the visual fields we inspected. A 0.5-ml aliquot of the PRP was added to each of four polypropylene reaction tubes (Falcon No. 2063), each of which contained 0.4 ml preincubation solution. This preincubation solution consisted of 0.1 ml phosphate buffer (1.0 mol/l, pH 7.4) in each reaction tube; two of the four tubes had 0.1 ml of pargyline (2.4 mmol/l, in 1 mmol/l HCl) to inhibit platelet MAO; distilled water was used to bring the preincubation solution in each tube to 0.4 ml. After addition of the PRP to the preincubation solution, the tubes were mixed with a Vortex apparatus. The duplicate values for each sample were averaged and the result utilized for all subsequent analyses. The duplicates always agreed within 5%.
Protein determination method The 4 ml of PRP from PP-1A and 1B or PP-8A and 8B was centrifuged at 2400 X g for 10 min to produce a platelet pellet. The supematant was discarded and the platelet pellet was washed twice with l-ml portion of ice cold saline. Distilled water, 0.4 ml, at ice bath temperature, was added to the pellet which was then sonicated for 10 set (setting 4, model S75-Heat Systems-Ultrasonics, Inc.) over ice. In an initial experiment, platelets were electronically counted in samples before and after sonication. The platelet count decreased by 98.2% ? 1.9% (N = 11) reflecting an almost complete disruption of the platelets by sonication. To assay MAO activity by the platelet protein method, four tubes were prepared for each sonicate in the manner described above with the exception that water was added to all tubes to a volume of 0.85 ml. To these reaction tubes, a 50-~1 aliquot of the platelet sonicate was added, following which the tubes were mixed with a Vortex@ apparatus. Both sets of reaction tubes were then preincubated at the same time in a shaking water bath at 37°C for 10 min in order for pargyline to inhibit platelet MAO. All the reaction tubes were then placed in an ice water bath for five min to temporarily halt the remaining MAO activity. One hundred ~1 of [14C]benzylamine (BA) (4.0 mmol/l, 600 &i/mmol, ICN Pharmaceuticals) was then quickly added to all the reaction tubes while in the ice bath, after which mixing was done with a Vortex@ apparatus. All the reaction tubes were then returned to the shaking water bath (37°C). After a 30-min incubation, all the tubes were again returned to the ice bath for 5 min so as to halt MAO activity, 0.3 ml of
18
3.0 mol/l HCI was then added and the tubes mixed with a Vortex@ apparatus to irreversibly inhibit the platelet MAO reaction. Three ml of heptane were added to all the reaction tubes which were then capped and mechanically shaken for 10 min to extract labelled reaction product. After shaking, the reaction tubes were centrifuged to break the emulsion and then placed in an acetone-dry ice bath to freeze the aqueous layer, allowing the heptane-product containing layer to be poured into counting vials. The amount of 14C-activity, representing the reaction products present in each heptane extraction, was determined with a liquid scintillation counter as disintegrations per min (dpm) per ml of reaction mixture. Recovery of the labelled reaction products was estimated by re-extracting samples with a second 3.0-ml portion of heptane. The total protein present in 50 ~1 of PS was determined by the method of Lowry et al. {421. Disintegrations per min due to platelet MAO activity were calculated as the difference between the average of tubes with and without pargyline [39]. Platelet MAO activity is reported as nmol of product per h per lo8 platelets or as nmol of product per h per mg protein for the platelet count and platelet protein methods respectively. Measures of reliability The reliabilities of either the platelet count assay method or the protein determination assay method were determined by calculating the coefficients of variation (C.V.) and intraclass correlations (I.C.C.) [43,44]. Re~~a~~lityof assay measures from same day duplicates Coefficients of variation and the I.C.C. were determined for the various assay measures obtained for the samples PC-1A and PC-1B and PCSA and PC8B from each subject. The measures which were evaluated were dpmiml reac tion volume, platelets/ml PRP and platelet MAO activity. A comparable series of C.V. and an I.C.C. were calculated using the same measures obtained with each subject’s PP-1A and PP-1B and PP-8A and PP-8B samples with the exception that protein/50 ~1 sonicate was used rather than platelets/ml PRP. The C.V. and I.C.C. formulated from these samples constitute the “same day pairs for day one” and are presented in Tables I-IV. Differences between the C.V. for the two methods were evaluated by means of a paired t-test. The C.V. and ICC, were also formulated for the same assay measures generated from the PC-8A and PC-8B, and from PP-8A and PP-8B for each subject and constitute the “same day pairs for day eight”. These results are also presented in Tables I-IV. Reliability of assay measures for assays done at seuen day i~teruals To compare the reliability of the measures obtained in blood samples from the same subject obtained on day one and on day eight, C.V. and I.C.C. were formulated from the means of the assay measures obtained for PC-1A and PC1B and for the means of the measures from PC-8A and PC-8B. The C.V. and I.C.C. were also formulated using the means of the assay measures described above from PP-1A and PP-lB, PP-8A and PP-8B. These C.V. and I.C.C. constitute the values referred to as “samples at seven day intervals” in Tables I-IV.
19
Results The mean platelet MAO activity using the platelet count assay was 6.89 t 1.66 units (mean kS.D.) for the 10 samples obtained on day one and 7.07 + 1.83 units for samples from the same 10 subjects obtained seven days later. The mean platelet MAO activity using the protein determination assay was 63.5 k 21.0 units for lo-samples from day one and 59.8 f 15.9 units for samples from the same 10 subjects seven days later, on day eight. These levels are within the normal limits for both assays in our laboratory. The Pearson correlation coefficients between the MAO activity determined by the platelet protein and platelet count methods using means of the pairs for each method from each subject for day 1 and day 8 were 0.775 and 0.762, respectively (p < 0.01). The mean C.V. for the same day pairs using the platelet count method from day one and day eight are 2.97% k 2.47% and 2.91 k 1.74%, respectively. Both values are significantly smaller 03 < 0.05) than the corresponding mean C.V. for day one and day eight, 5.62% + 4.35% and 5.43% + 3.89%, respectively, which were obtained with the protein determination method (Table I). The mean C.V. of the two methods for the samples obtained at an interval of seven days were not significantly different, in part because of the large variance in the platelet protein samples. The mean C.V. for the platelet count method was less than half that of the protein method. There was as much as a 30% discrepancy in the samples obtained at a ‘I-day interval with the platelet protein method whereas no subject had MAO levels which differed by more than 11% with the platelet count method. The I.C.C. for all day one or all day eight pairs was highly significant and virtually identical for both methods (0.979 and 0.981 for the platelet count method and 0.970 and 0.952 for the platelet protein method) (Table II). The I.C.C. based on the mean of split duplicates from day one and day eight of the study was considerably better for the platelet count method, 0.941, than for samples assayed using the protein determination method, 0.852 (Table II). However, both were significant at the 0.01 level. To further identify the possible contribution of variance in both assays, we
TABLE I PLATELET MAO ACTIVITY DETERMINATIONS (C.V.) IN TEN MALE SUBJECTS
AND
THEIR
COEFFICIENTS
Platelet assay
Same day pairs Mean day 1 Mean day 8
OF VARIATION
Protein *ssay
Activity *
C.V. (96)
6.89 + 1.66 7.07 t 1.83
2.97 f 2.47 2.91 * 1.74
6.98 r 1.69
4.25 f 3.63
9 5
Activity **
C.V. (%)
63.5 + 21.0 59.8 f 15.9
5.62 + 4.35 5.43 r 3.89
61.6 * 18.2
8.93
Samples at a ‘I-day interval Using mean values for day 1 and day 8
* nmol product/lo8 platelets (mean f S.D.). ** nmol product/h/mg protein (mean f S.D.). 5 Coefficients of variation are significantly different, P < 0.05 by paired t-test.
? 9.76
5 8
20 TABLE
II
INTRACLASS
CORRELATIONS
Same day pairs Day 1 Day 8 Samples at a 7-day interval * nmol product/h/lo8 * * nmol product/h/mg ***p < 0.01.
FOR PLATELET
MAO ACTIVITY
IN TEN MALE
Platelet assay *
Protein assay * *
0.979 0.981 0.941
0.970 0.952 0.852
*** *** ***
SUBJECTS
*** *** ***
platelets. protein.
examined the reliability of the various component measures of the platelet MAO activity. In Tables III and IV, the C.V. and I.C.C. of the dpm/ml of reaction volume, the platelets/ml of PRP and the protein/50 ~1 of sonicate are compared. The C.V. for dpm/ml of reaction volume were not significantly different for day one pairs, for day eight pairs or for samples obtained at a seven day interval for the two assay methods (Table III). The C.V. for platelets/ml of PRP and protein/50 ~1 of sonicate of day one and day eight pairs (Table III) are significantly different: the platelet counts from both days are significantly more reliable than the corresponding protein determinations (p < 0.025). There was no significant difference between C.V. of platelet counts or protein determinations in samples obtained at the seven day interval (Table III). The data of Table IV indicates that the I.C.C. for dpm/ml of reaction volume and for platelets/ml were significant at p < 0.01 while the I.C.C. for protein/50 1_r1 of sonicate determinations were significant at p < 0.05. We also compared the dpm/h/ml of reaction volume for the two methods.
TABLE
III
COEFFICIENTS
OF VARIATION
FOR COMPONENT
MEASURES
OF PLATELET
C.V. (%) Platelet assay
Protein assay
dpm/ml of reaction volume Same day pairs Mean day 1 Mean day 8 Samples at a 7-day interval
5.68 ? 4.89 4.06 k 3.39 7.35 k 7.60
6.16 6.61 7.98
Platelets/ml of PRP Same day pairs Mean day 1 Mean day 8 Samples at a ‘T-day interval
2.01 2.53 9.10
-
Protein/50 ~1 of sonicate Same day pairs Mean day 1 Mean day 8 Samples at a ‘T-day interval
-
* P < 0.025
when compared
? 2.12 * + 1.14 * ? 6.66
to same day split duplicates
8.46 8.06 9.96
2 4.23 f 5.91 it 6.37
+ 6.67 + 3.26 + 5.63
of protein150
~1 of sonicate.
MAO
ACTIVITY
21 TABLE
IV
INTRACLASS
CORRELATIONS
FOR
COMPONENT
MEASURES
OF PLATELET
MAO ACTIVITY
I.C.C. Platelet
assay
Protein
dpm/ml of reaction volume Same day pairs Mean day 1 Mean day 8 Samples at a 7-day interval
0.967 0.974 0.853
* * *
0.964 0.942 0.939
Platelets/ml of PRP Same day pairs Mean day 1 Mean day 6 Samples at a 7-day interval
0.986 0.989 0.776
* * *
-
Protein/50 ~1 of sonicate Same day pairs Mean day 1 Mean day 8 Samples at a ‘I-day interval
-
0.590 0.835 0.596
assay
* * *
** * **
* P < 0.01. * * P < 0.05.
This comparison is meaningful because the reaction volume and the reagent concentrations are the same in both assay methods while the quantity of platelet MAO protein in 50 ~1 of sonicate used in the platelet protein assay should be the same as the quantity of platelet MAO protein contained in 0.5 ml PRP. We found that the mean platelet MAO activity in the sonicate was 28.5% 2 12.7% less than that in the PRP used for the platelet count assay in samples from day one and 32.9% ?r 4.7% less in determinations from day eight samples. Combining the results of both days, the mean dpm/h/ml of reaction volume was 30.8% + 7.5% less for the protein assay than for the platelet count assay (p < 0.001, paired t-test). Discussion The results reported here indicate that the reliabilities of the platelet MAO activity and the component measures of that activity in same day paired samples for the platelet count assay method tend to be greater than that of the protein assay method. While the absolute differences in assay reliabilities are not great, every type of comparison we employed reflected a greater reliability for the platelet count method than for the protein method. Our results should be considered a minimum difference between the two methods because of the great care which we applied to all aspects of these determinations, including assaying each sample in duplicate. A probable cause for the lesser reliability of the protein assay versus the platelet count assay is a variable contribution of protein from red and white blood cells plus other noncellular proteins and aromatic amino acids which are measured along with platelet protein [ 19,30,39,40]. This conclusion is partially
22
supported by our data which shows no significant differences in reliability between the dpm/ml reaction volume generated by either assay method while the variability of the total protein determination was substantially greater than the variability of the platelet counts for same day duplicates (Tables III and IV). Since we found no WBCs or RBCs in the PRP which we examined in the microscope, we can assume that non-cellular protein is the source of the contaminating protein. Our results also indicate that there are significant differences in the MAO activity which can be obtained from the same PRP by the two methods, The loss of approximately 30% of MAO activity as a result of the sonication which is employed in the protein determination method suggests that the platelet count method could provide a more valid estimate of platelet MAO activity than the protein determination method. The majority of studies of platelet MAO activity in schizophrenics have employed the somewhat less reliable platelet protein method [l-10,14-24,28,29, 31--33,35,36]. Of these 28 studies, 16 reported decreased platelet MAO in schizophrenics. Seven of nine studies with the platelet count method reported decreased platelet MAO activity (Chi square = 0.512). Thus, there is little indication that the studies which utilize the platelet count method demonstrate a lower platelet MAO activity in schizophrenics more frequently than studies which utilize the platelet protein method. If done with care, the platelet protein method can achieve acceptable reliability but is is very easy for large errors to occur because of the effects of sonication and the possible trapping of nonplatelet protein in the pellet [19]. In our hands, the protein determination showed an average 8.3% i 3.8% variation in same day paired samples. The average C.V. for electronic platelet counting was 2.2% ?: 1.2% for same day pairs. Therefore, we believe the platelet count method for measuring platelet MAO activity, with electronic platelet counting, is the preferred method for determining platelet MAO activity. Our current investigation of platelet MAO activity in schizophrenics utilizing the platelet count method suggests that there is a subgroup of schizophrenics, characterized by hallucinations, who have signific~tly lower platelet MAO activity than non-hallucinating schizophrenics and normal controls (Meltzer, H.Y., Arora, R. and Jackman, H., in preparation). This is in agreement with previous reports of Schildkraut et al. [ 181 and our laboratory [24] and suggests that patient population may be a significant factor explaining differences between studies. Acknowledgements This study was supported in part by USPHS 25 116 and MHCRC 30 938. H.Y.M. is recipient of TJSPHS RCSA MH 47 808. References 1 Murphy. D. and Wyatt, R. (1972) Nature 238, 225-226 2 Wyatt, R., Murphy, D., Belmaker. R., Cohen. S., Donnelly, 916filS
C. and Pollin, W. (1973)
Science 179.
23 3 Friedman, E., Shopsin, B., Sathononthan, G. and Gershon. S. (1974) Am. J. Psychiatr. 131, 13921394 4 Meltsex, H.Y. and Stahl, S. (1974) Res. Common. Chem. PathoI. Pharmacd. 7.419431 5 Nies. A., Robinson, D., Harris, L. and Lamborn, K. (1974) in Advances in Biochemical Psychopharmacology (Usdin, E.. ed.), pp. 59-70, Raven Press, New York 6 Owen, F., Ridges, P. and Cookson, I. (1973) Acta Genet. Med. GemeIIol. (Roma) 23.371-376 ‘7 Carpenter. W.. Murphy, D. and Wyatt, R. (1975) Am. J. Psychiatr. 132.438441 8 Cookson, I., Owen, F. and Ridges, A. (1975) Psychol. Med. 5. 314-316 9 Shaskan, E. and Becker, R. (1975) Nature 253.659-660 10 Takahashi, S., Yamane. H. and Toni, N. (19753 FolIa Psychiatr. Neurol. Jap. 29.207-214 11 ZeIIer, E., Bashes, B., Davis, J. and Thorner, M. (1975) Lancet i, 1385 12 Belmaker. R., Ebstein, R., Rimon, R., Wyatt. R. and Murphy, D. (1976) Acta Psychiatr. Stand. 54. 67-72 13 Belmaker, R., Ebbesen, K.. Ehstein, R. and Rimon, R. (1976) Br. J. Psycbiatr. 129, 227-232 14 Brockington, I., Crow, T., Johnstone, E. and Owen, F. (1976) in Monoamine Oxidase and its inhibitors (Ciba Inc., ed.), pp. 353-362, Elsevier, New York 15 Domino, E. and Khanna, S. (1976) Am. J. Psychiatr. 133, 323-326 16 Murphy. D., Donnelly, C.. Miller, L. and Wyatt, R. (1976) Arch. Gen. Psychiatr. 33.1377-1381 17 Owen. F., Bourne, R., Crows, T., Johnstone, E., Bailey, A. and Her&ion, H. (1976) Arch. Gen. PSYchiatr. 33.1370-1313 18 SchiIdkraut, I., Herzog, J., Orsuiak, P., Edelman. S.. Stein, H. and Frazier, S. (1976) Am. J. Psychiatr. 133,438--440 19 White, H., McLeod. M. and Davidson. J. (1976) Am. J. Psychiatr. 133,1191-1193 20 Wyatt, R. and Murphy. D. (1976) Schiz. Bull. 2.77-89 21 Becker, R. and Shaskan, E. (1977) Am. J. Psychiatr. 134,512--517 22 Belmaker, R., Galon, A., Perez, L. and Ebstein, R. (1977) Br. 3. Psychiatr. 131. 551-552 23 Joseph, M., Owen, F.. Baker, H. and Boume, R. (1977) PsychoI. Med. 7.159-162 24 Meltzer, H.. Pscheidt, G., Goode, D.. PiyakaImaIa, S. and Dichtel, M. (1377) presented at the American Psychiatric Association Meeting, Toronto. Canada, May, 1977 25 Murphy, D.. Bebnaker, R., Carpenter, W. and Wyatt, R. (1977) Br. J. Psychiatr. 130.151-158 26 Sullivan, J., StanfieId,C. and Da&is, C. (1977) Am. J. Psychiatr. 134.1098-1103 27 Berger, P.. Ginsberg, R., Barchas, J., Murphy, D. and Wyatt, R. (1978) Am. J. Psychiatr. 135.9549 28 Friedhoff, A., Miller, J. and Weisenfreund, J. (1978) Am. J. Psyehiatr. 135.952-955 29 Groshong. R., Baldesstini, R., Gibson, D.. Lipinski, J., AxeIrod, D. and Pope, H. (19781 Arch. Gen. Psychiatr. 135,1198-1205 30 Potkin, S,. Cannon, H., Murphy, D. and Wyatt, R. (19781 New Eng. J. Med. 298.61-66 31 Schildkraut, J.. Orsulak. P.. Schatzberg, A., Cole, J., Gudeman. J. and Rhode, W. (1978) Am. J. PSYchiatr. 135,110-112 32 SuIIivan, J., Cavenar, J., Stanfield, C. and Hammett, E. (1978) Am. J. Psychiatr. 135. 597-698 33 Valkenberg, C. and Crowe, R. (1978) New Eng. J. Med. 298, 1150 34 Van Kammen, I).. Marder. S., Murphy, D. and Burmey, W. (19781 Am. 3. Psychiatr. 35.600-605 35 Berrettini, W., Prozialeck, W. and Vogel, W. (1978) Arch. Gen. Psychiatr. 35, 600-605 36 Berrettini, W. and Vogel, W. (1978) Am. J. Psychiatr. 135.605-607 37 SuIIivan. J., MaItbie. A., Cavenar, J.. Hammett, E. and Lister, P. (1978) New Eng. J. Med. 298,115l 38 Wyatt, R., PotkIn. S.. GiIbin, J. and Murphy, D. (1978) in Psychopharmacology: A Generation of Progress (Lipton, M., et al., eds.), pp. 1093-1095, Raven Press, New York 39 Murphy, D., Wright, C., Buchsbaum, M., Nichols. A., Costa, J. and Wyatt, R. (1976) Biochem. Med. 16.254-265 40 Meltzer. H.Y. (1978) in The Schizophrenia Syndrome (BeIIak, L., ed.), Grune and Stratton, New York, in press 41 Youdim, M.B.H. and Holzbauer. M. (1976) J. Neurrd Trans. 38.193-229 42 Lowry, O.H., Rosebrough, N.J., Farr. A.L. and RandaIl. R.J. (1951) J. Biol. Chem. 193,265-275 43 Snedecor, G. and Cochran, W. (1967) Statistical Methods, 6th edn., Iowa State University Press, Ames, Iowa 44 Bartko, J. and Carpenter, W. (1076) J. Nerv. Ment. Dis. 163.307-317
