‘ “.1. J ‘_A ‘5 ,1] l|\\ ' '3’ ' I ‘ saw,» OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MRTERIALS: Place in book return to remove charge from circuiation records GENETIC AND BIOCHEMICAL STUDIES OF THE QUANTITATIVE VARIATION OF NUCLEOSIDE TRIPHOSPHATE PYROPHOSPHOHYDROLASE IN HUMAN ERYTHROCYTES by Steven Andrew Fuller A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY. College of Natural Science Genetics Program 1980 cvnbing/ ABSTRACT GENETIC AND BIOCHEMICAL STUDIES OF THE QUANTITATIVE VARIATION OF NUCLEOSIDE TRIPHOSPHATE PYROPHOSPHOHYDROLASE IN HUMAN ERYTHROCYTES BY Steven Andrew Fuller Nucleoside triphosphate pyrophosphohydrolase (NTPH) specific activity exhibits extreme variation in the red cells of the human population. Previous data have shown that this variation in NTPH activity among individuals is mirrored in granulocytes, lymphocytes, and platelets as well. Studies exploring the molecular basis of the variation of NTPH specific activity have presented evidence that this variation is not due to the presence of intracellular effectors or to differences in the Km for its substrate ITP. Further data are here presented which examine the biochemical and genetic aspects of NTPH variability in human red cells. A purification scheme for human red cell NTPH is presented which results in a nearly homogeneous preparation of theeenzyme. Estimation of native and subunit molecular weights of NTPH by gel filtration, sucrose density gradient centrifugation, and sodium dodecyl sulfate-polyacrylamide gel electrophoresis suggest that NTPH is a dimer composed of identical 40,000 dalton subunits. The pI of human red cell NTPH as determined by isoelectric focusing is about 4.7. Confirming earlier experience, data are presented that NTPH requires the presence of sulfhydryl compounds, Steven Andrew Fuller in particular dithiothreitol for maximal activity. Reagents that react with SH groups of enzymes, such as N-ethylmaleimide, iodoacetamide, and p-hydroxymercuribenzoate, inhibit NTPH activity. A survey of NTPH in the red cells of a random Caucasian population reveal clearly distinct high and low specific activity modes, with approximately 18% of the population to be found in the low activity mode. Further studies of the biochemical basis for the variation in NTPH activity show that NTPH of high and low activity individuals does not differ in the utilization of alternate substrates dITP and XTP. Likewise, no differences were observed in the in give stability of the enzyme as measured in red cell populations separated by cell age. Thermostability experiments, however, provide evidence for two distinct phenotypes in the low activity mode and a third phenotype in the high activity mode. NTPH activity in cells with low specific activity was less heat stable than that in cells of high specific activity. Family studies were undertaken to examine the genetic basis of the variation of human red cell NTPH specific activity. Red blood cells from members of 57 families were analyzed for NTPH activity. In conjunction with the differences observed in thermostability, pedigree analysis presents evidence that NTPH variability cannot be attributed to a simple one gene-two allele system of inheritance. It is QbCVClL HLLULCW L' ULLCJ. proposed that a genetic model involving three alleles at one gene locus can account for the genetic and biochemical observations presented here. TO my wife, Nancy: my son, Jeffrey: and our families. ii ACKNOWLEDGEMENTS I would like to express my appreciation to Dr. A.J. Morris for his guidance, encouragement, and friendship during my time in his laboratory. I would also like to thank the members of my guidance committee: Dr. A. Ellingboe, Dr. F. Rottman, and Dr. J. Asher. Dr. Asher was particularly helpful in discussions. I thank my fellow graduate students, V. Verhoef, W. Chaney, C. Vary, and H. Hershey for their helpful discussions and assistance. iii TABLE OF CONTENTS List of Tables List of Figures I. Introduction II. Literature Review A. Studies of Nucleoside Triphosphate B. Pyrophosphohydrolase Genetic Mechanisms of Quantitative Variation of Enzyme Activity 1. Biochemical bases of enzyme variation 2. Hypoxanthine-guanine phosphoribo- syltransferase Glucose-G-phosphate dehydrogenase Pyruvate kinase Red cell acid phosphatase Peptidase A Catehhol-0-methyltransferase Galactokinase ooqoxcnhw o o o o o 0 III. Reagents, Materials and Methods A. Reagents l. Commerical 2. Recrystallization of acrylamide and N,N'-methylene-bis-acrylamide Materials 1. Commerical 2. Special preparation of glass tubes for polyacrylamide gel electrophoresis 3. Biological-blood samples a. population survey b. Families selected for genetic analysis c. Other families obtained d. Psychiatric patient samples e. miscellaneous samples used for study iv y... Page vii viii l7 17 22 25 28 29 30 31 32 34 34 34 36 36 36 37 37 37 37 38 38 38 IV. C. l. 2. Results A. Methods Analyticsl a/ NTPH, phosphate, hemoglobin, and protein analyses b. Sucrose density gradient centrifugation c. Isoelectrié focusing in sucrose density gradients d. Isoelectrid focusing in polyacrylamide gels e. Polyacrylamide gel electro- phoresis f. Sodium dodecyl sulfate-poly- acrylamide gel electrophoresis g. Solubilization of poly- acrylamide gels h. Protein staining of poly- acrylamide gels i. Staining of polyacrylamide gels for NTPH activity j. Inhibition studies with N- ethylmaleimide, iodo- acetamide, and p-hydroxymercuri- benzoate k. Red cell separation by age 1. Thermostability experiments m. Computer analysis of data General Procedures a. Preparation of hemolysates b. Purification of human erythrocyte NTPH Physical Studies of Human Red Cell NTPH Determination of isoelectric point by isoelectric focusing in sucrose density gradients l. a. b. Determination of pI of NTPH using human red cell hemolysates Determination of pI of NTPH using partially purified human red cell NTPH Further purification of human red cell NTPH Molecular weight estimations of humanrred cell NTPH a. b. Gel filtration of partially purified NTPH Molecular weight estimation of human red cell NTPH by sucrose density gradient centrifugation ‘11 v 39 39 39 41 42 43 44 45 47 49 49 50 51 52 52 52 53 56 56 56 S6 59 60 60 c. Subunit molecular weight estimation of human red cell NTPH by sodium dodecyl sulfate- 66 polyacrylamide gel electrophoresis 4. Studies of the thiol requirements of human red cell NTPH B. Population Surveys of NTPH Activity in Human Red Blood Cells 1. NTPH activity in a normal population 2. NTPH activity in the red cells of patients with paranoid schizophrenia and schizophrenia C. Studies of Biochemical Parameters of NTPH Variation 69 80 80 86 86 1. Isoelectric focusing in polyacrylamide gels 2. Substrate specificities of red cell hemolysates 3. Red cell separation by age 4. Thermostability of NTPH D. Family Studies of NTPH Variation V. Discussion A. Physical Studies of NTPH B. Analysis of the Genetic Variation of Human Red Cell NTPH Activity C. The Metabolic Role of NTPH and Consequences of Its Genetic Variation List of References vi 86 88 89 95 104 135 135 137 148 155 Table Table Table Table Table Table Table Table Table Table Table Table 10 11 12 LIST OF TABLES Inhibition of NTPH Activity by Sulfhydryl—Specific Reagents Effect of Selected Sulfhydryl Compounds Upon NTPH Stability NTPH Activity in the Red Cells of Psychiatric Patients NTPH Activity with Alternate Substrates Red Cell Separation by Density (Age) NTPH Specific Activity and Thermostability at 65°C Half-life of NTPH at 65°C Test Hypotheses for Inheritance of NTPH Variability Pedigree Data of Families of Figures 17 and 18 HLA and ABC Blood Typing of Family T Members Codominant Hypothesis of Inheritance of NTPH Variability A Model of the Inheritance of NTPH Specific Activity vii Page 71 73 87 94 100 103 107 120 124 124 131 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 10 11 12 13 14 15 LIST OF FIGURES Isoelectrié Focusing of NTPH of a Red Cell Hemolysate Isoelectric Focusing of Partially Purified Red Cell NTPH Polyacrylamide Gel Electrophoresis of NTPH Purification Fractions Sephadex G-100 Gel Filtration of Partially Purified NTPH Sucrose Density Gradient Centrifugation of NTPH Determination of S Value and Molecular Weight of NTPH SDS-Polyacrylamide Gel Electrophoresis of Purified Human NTPH Determination of Subunit Molecular Weight of NTPH NTPH Activity During Incubation in Presence or Absence of Sulfhydryl Compounds Reactivation of Inactivated NTPH by Dithiothreitol A Random Survey of NTPH Specific Activity in the Red Cells of a Caucasian Population Normal Distributions Fit to NTPH Population Survey Red Cell Separation by Density (Age) Thermostability at 65°C of Hemolysates of Two Subjects Semilog Plot of Data of Figure 14 viii Page 57 58 61 63 64 65 67 68 75 78 82 85 92 96 98 Figure Figure Figure Figure Figure Figure Figure Figure 16 17 18 19 20 21 22 23 Figurer24 Figure 25 NTPH Specific Activity and Thermostability at 65°C NTPH Specific Activity of Randomly Selected Families NTPH Specific Activity of Non- Randomly Selected Families NTPH Specific Activity of Parents of Figure 17 NTPH Specific Activity of Family Members of Figure 17 NTPH Specific Activity of Males of Figure 17 NTPH Specific Activity of Females of Figure 17 NTPH Specific Activity and Thermostability in Two Families NTPH Activity Distributions of Offspring of Various Parental Pairs Proposed Genotypes of Members of Nine Families ix 101 109 113 115 116 H7 118 126 128 146 I. Introduction A unique pyrophosphohydrolase activity was detected in this laboratory during investigation of in_zit£g globin synthesis in rabbit reticulocyte lysates when a GTP hydrolase activity was observed. This activity was subsequently purified over 2000-fold from rabbit reticulocytes and characterized as a nucleoside triphosphate pyrophosphohydrolase (NTPH, EC 3.6.1.19) (1). This enzyme catalyzes the general reaction, NTP + H20 + NMP + PPi, where NTP and NMP refer to nucleoside triphosphate and monophosphate, respectively. ITP and dITP are the most readily used substrates, while XTP is used at 71%, and UTP and GTP are used at about 10% of the activity with ITP. ATP, IDP, and IMP are not hydrolyzed by NTPH. The purified enzyme has optimal activity at pH 9.75 in the presence of 10 mM MgC12, 0.5 mM ITP and a sulfhydryl compound such as DTT. The estimated molecular weight was 37,000 daltons. Subsequent studies included a survey of NTPH in various tissues of the rabbit which revealed that the enzyme is present in all of eleven tissues examined, having highest activity per cell in the brain and liver (2). The erythro- cytes have the lowest activity per cell of all tissues examined. Rabbit liver NTPH was purified 660-fold and was found to exhibit the same properties as the reticulocyte NTPH. An examination of electrophoretic mobility of NTPH in polyacrylamide gels showed no evidence of isozymes. NTPH activity was also found in the red blood cells of a number of other species including man. The most striking feature of NTPH levels in human red cells was the marked variation from one individual to another. While the NTPH specific activity (nmoles ITP cleaved/20 minutes/mg hemoglobin at 37°C) of each individual remains constant over long periods of sampling, specific activity varies over a range of from 0 to over 100 between individuals. That NTPH variability in human red cells may be an inherited trait was first proposed in an indirect manner by Vanderheiden (3). He has observed elevated levels of ITP in blood samples of two siblings and suggested that this trait was inherited. Further investigation found elevated ITP in 7 of 6000 subjects (4) which was correlated with deficiency of an "ITPase". A population survey and studies of a small number of families by Vanderheiden led him to conclude that "ITPase" deficiency was inherited as an autosomal recessive trait. The "ITPhsed" activity measured by Vanderheiden was actually the result of a coupled reaction between NTPH and endogenous inorganic pyrophosphatase. Because of uncertainities of Vanderheiden's data, particularly the use of subOptimal enzyme assay conditions, improper selection of subjects for population data, and limited family data, it was of interest to this laboratory to further examine NTPH variability in human red cells. A collaborative study with J. L. Henderson's laboratory at the University of Alberta (Edmonton, Alberta, Canada) examined the relationship between NTPH activity of erythrocytes and the ability of the erythrocytes to accumulate [14C] ITP i_n_ 33559 from added [14¢] hypoxanthine (5). The relationship between these two parameters fits closely a theoretical relationship predicted by employing Michealis-Menten kinetics for the relationship of an enzyme activity with the con- centration of the substrate. The biochemical basis of NTPH variability has previously been examined by two methods. First, a mixing study was carried out using hemolysates of individuals differing widely in NTPH specific activity. The strictly additive nature of activity found in the mixtures provided no evidence for the existence of intracellular inhibitors or activators of NTPH activity. A similar study mixing hemolysates with partially purified human red cell NTPH led to the same conclusions (5). Secondly, the Km for ITP of NTPH of hemolysates was examined in selected individuals (5). Results of the Km determinations showed no pattern of variation associated with different levels of NTPH activity in the hemolysates, although.measurement of Km's in individuals with very low levels of NTPH activity was not possible for technical reasons. Other studies have shown that NTPH activity of an individual's red blood cells is directly correlated with the NTPH activity of that individual's granulocytes, lymphocytes, and platelets (6). These experiments support the contention that NTPH activity is inherited in a defined manner in all tissues. Data presented here expand the genetic and biochemical studies NTPH in human red blood cells. Further work on the isolation of human red cell NTPH is presented that results in the highest degree of purification yet achieved. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of purified NTPH reveals the presence of two polypeptides. The molecular weights of these polypeptides are 26,000 and 40,000 daltons. The 26,000 dalton polypeptide has been eliminated in more recent preparations. The subunit molecular weight of NTPH is thus proposed to be approximately 40,000 daltons. An isoelectric focusing study shows that the pI of human red cell NTPH, as determined in lysates and partially purified enzyme, is about 4.7. A recent report of NTPH purification by Vanderheiden stated that NTPH which has been purified some 2000-fold does not require sulfhydryl reagents for activity or stability (7). Since it is the experience of this labOratory that NTPH does indeed require the presence of sulfhydryl reagents, further experimentation was carried out in this area. The results confirmed that sulfhydryl compounds, in particular dithiothreitol, were required for maximal NTPH activity and for stability in purified preparations. Reagents that ‘\ J react with SH groups of enzymes such as N-ethylmaleimide (NEM), iodoacetamide (IAA), and p-hydroxymercuribenzoate (pHMB) were incubated with lysates and purified NTPH. NEM and IAA concentrations of .05 M inhibited NTPH activity 100% and 71%, respectively, while pHMB at a concentration of .01 M provided total inhibition of activity. A survey of NTPH activity in a random Caucasian population revealed at least a bimodal distribution. Approximately 18% of the population had specific NTPH activity below 25. The "high" mode ranged in specific activity from 25 to about 120. A possible third mode con- sisted of those individuals with NTPH activity ranging from 0 - 5. Studies of the biochemical basis of the variability of red cell NTPH activity in the human population were carried out in an attempt to correlate differences in biochemical properties of NTPH with the observed variation of NTPH activity. Isoelectric focusing of NTPH in polyacrylamide gels provided no evidence of isozymes when the gels were stained for NTPH activity. The only differences in NTPH which.were observed were in the degree of staining intensity which was proportional to the specific activity of the NTPH lysate used. The substrate specificities of NTPH from high and low activity individuals were also examined using.XTPh,dITP and GTP. it i m\~ ‘I 51 All individuals utilized these alternate substrates in the same proportion when ITP hydrolysis was taken as the reference for comparison. The in_vivg stability of NTPH was examined by the separation of red cell populations according to their density (age). The activity of the oldest cells was compared to that of the youngest cells. While NTPH activity declines with the age of the red cells, no differences in the rate of NTPH loss with age, associated with NTPH specific activity, were observed. These experiments provided evidence that low NTPH activity was not due to an increased rate of degradation, or lability of the enzyme during the life-time of the erythrocyte. Finally, the thermostability of NTPH in lysates of individuals with a wide range of NTPH specific activities was examined. This parameter provided the first evidence of biochemical differences in red cell NTPH of individuals with markedly different specific activity values of NTPH. Three phenotypes were distinguished by thermostability. Individuals in the high specific activity NTPH range (> 25 specific activity) had enzyme which was more heat stable than those in the 5 — 25 range of specific activity. Individuals with NTPH specific activity below 5 have enzyme that is significantly more thermolabile than the other two groups. These results provide evidence dér differences in the amino acid sequence of NTPH of red cells of individuals from the above groups. «r (.I' An Family studies were carried out involving members of 57 families. In combination with the differences in thermo- stability, evidence is presented using pedigree analysis that NTPH variability in human red cells cannot be attributed to a simple one gene-~two allele system of inheritance, but rather that a more complex system involving multiple alleles must be involked to explain these data. i J II. Literature Review A. Studies of Nucleoside Triphosphate Pyrophosphohydrolase In the course of studying nucleoside triphosphate metabolism in rabbit reticulocytes, this laboratory detected a unique pyrophosphohydrolase activity. This nucleoside triphosphate pyrophosphohydrolase (NTPH, EC 3.6.1.19) was subsequently purified over 2000-fold from rabbit reticulocytes and characterized (l). The reaction catalyzed by NTPH was shown to be NTP + NMP + PPi, where NTP and NMP refer to nucleoside triphosphate and monophosphate, respectively. ITP and dITP were the most readily used substrates. Other nucleoside triphosphates, XTP, UTP and GTP were hydrolyzed at rates of 71%, 12% and 10% that of ITP, respectively. ATP, IDP and IMP were not hydrolyzed by NTPH. The purified NTPH had a pH optimum of 9.75 with B-alanine buffer, required divalent cations, having optimal activity with 10 mmngClz, and required the presence of a sulfhydryl compound. Monovalent cations were inhibitory of ITP hydrolysis by NTPH, while of the nucleotides examined, IDP had by far the greatest inhibitory effect. The estimated molecular weight was 37,000 as determined by sucrose density gradient centrifugation. Data from gel filtration which gave a molecular weight estimate of 70,000 suggested the possibility that NTPH was a dimer composed of 37,000 daltdn subunits. sh sag). NTPH activity was also found in the red blood cells of a number of species including man, mice, rats, cattle, sheep, horses and snakes. The most striking feature of NTPH levels in human red cells was the marked variation from one individual to another, ranging from undetectable levels to a specific activity of 120. (1 unit of NTPH activity is that amount of enzyme activity which cleaves l nanomole of ITP in 20 minutes of incubation at 37°C. Specific activity is expressed as units per mg hemoglobin.) The activity level of each individual remains constant over long periods of sampling. Investigation of other cell types revealed that NTPH activity was present in all eleven tissues of the rabbit which were examined (2). Brain and liver had approximately 66 and 42 times the activity of erythrocytes, respectively, on a per cell basis. The erythrocytes had the lowest activity per cell of all tissues examined. NTPH was also purified from rabbit liver (2). The substrate specificity, pH optimum, electrophoretic mobility in polyacrylamide gels, and apparent molecular weight of rabbit liver NTPH were essentially identical to those properties of the rabbit red cell enzyme. Subsequent partial purification of human erythrocyte NTPH some IZOD-‘-~ fold (8) revealed properties quite similar to those of the rabbit. Certain differences eXisted, including that of a lower utilization of GTP and an apparently lower degree 0f stability of the human enzyme in purified preparations. 10 Other laboratories have been involved in the study of NTPH in red cells, although before the report of Chern §E_al. (l) the enzyme was thought to be a phosphatase rather than a pyrophosphohydrolase. Liakopoulou and Alivasatos (9) described such an "ITPase" in human erythrocytes. IDP and inorganic phosphate were thought to be the reaction products. Vanderheiden (4) also examined a human red cell "ITPase" via a population survey and pedigrees, while Hershko‘gt El; (10) studied an "ITPase" in rabbit red cells. "ITPase" activity was most likely observed as the coupled reaction of NTPH with endogenous inorganic pyrophosphatase (11). Reports concurrent with and following the paper of Chern gt_al; (1), confirmed the pyrophosphorylytic cleavage of nucleoside triphosphates by NTPH in rabbit (12) and human (13) red cells, although some of the properties of NTPH examined were different than those found in this laboratory. In the studies reported in these latter publications the rabbit NTPH preparation was purified only loo—fold, while the human red cell relative purification was unspecified. A recent report (7) of a purification of human NTPH some 2000-fold stated that NTPH exhibited no requirement for sulfhydryl compounds for enzyme activity. However, the purification steps were all carried out in the presence of mM 2+mercaptoethanol. Previous evidence from this laboratory and evidence to be presented here indicate that 11 sulfhydryl compounds are required for stability and maximal activity of NTPH. A study of NTPH activity in rat tissue (14) corroborated the evidence of Wang and Morris (2) in the rabbit that NTPH is ubiquitous in cytosol fractions of tissue extracts. NTPH activity/mg protein was found to be highest in the adrenals and thymus, two tissues not previously studied. Brain, liver and kidney were next highest, while erythro- cytes were found to be lowest in activity, the same order found in the rabbit, although Wang and Morris measured NTPH activity on a per cell basis. While this laboratory had observed the marked variation of NTPH levels in human erythrocytes among individuals, the idea that NTPH variability was an inherited trait was first put forward in an indirect manner by Vanderheiden (3). He detected elevated ITP levels in red cell lysates of individuals of one family. In an expanded study (4) seven of 6000 blood samples from a mixed population also revealed the presence of high ITP concentrations. It was hypothesized that a deficienty of "ITPase" activity was the cause of the elevated ITP levels. Erythrocytes from an unspecified number of individuals with "low", "intermediate", and "high" ITP concentrations in their cells were incubated with I14CU inosine. The ratios of cpm found in [14C] IMP were .001 - .003, .01 - .03 and .14 — .21, respectively. Three distinct levels of "ITPase" were inversely correlated with Jim () (2 f) 12 the [14C] ITP/[14C] IMP ratios. Those individuals with the highest incorporation of radioactivity into ITP had low "ITPase" activity. The same report gives the results of a population survey of ITTPase" activity in human red cells, revealing a bimodal distribution of activities, and some pedigrees which form the basis of a proposed autosomal codominant mode of inheritance. These data suffer from un- certainties, however, as the analysis of "ITPase" was carried out under suboptimum.conditions and there existed a possible limitation of phOsphate formation from inorganic pyrophosphate by available endogenous pyrophosphatase. Also, the subjects for the population survey were drawn from a mixture of racial groups and many of the pedigrees presented are lacking one parent, which makes analysis of data unreliable. The relationship between NTPH activity and ITP levels was examined in a somewhat different manner by Soder §t_al, (5) in a collaboration between this lab and that of J. F. Henderson of the University of Alberta, Edmonton, Alberta, Canada. The rate of [14C] ITP accumulation in human red cells from.added I14C] hypoxanthine was correlated with the specific activity of NTPH in thOse cells. In contrast to vanderheiden's finding, this relationship reVealed a continuum of NTPH specific activity and [14C] ITP accumulation valaies, not being broken up in three distinct groups, as Vanderheiden had reported. These data followed closely the 13 theoretical relationship between enzyme activity and steady state substrate concentration as calculated from the Michealis-Menten equation using percent [14C] ITP as the measure of Vmax‘ These data are most readily interpreted as indicating that the inherited level of NTPH in the red cell determines the intracellular ITP concentration, rather than variations in the biosynthetic rate of ITP. A population survey reported here indicates the existence of two and possibly three phenotypes of NTPH activity levels. A distinct bimodality exists with one group consisting of 0 - 25 units NTPH activity and the other, 25 - 120 units. In the study discussed above, [14C] ITP accumulation increases dramatically as activity decreases below an NTPH level of about 30. The variation of NTPH activity among individuals is not restricted to the red cells. A strong positive correlation was found to exist between the NTPH activity in the red cells of individuals and that in their granulocytes, lymphocytes, and platelets (6). The activity per cell in granylocytes was approximately 8 - 10 times that in red cells, while the activity per cell in lymphocytes was more than 20 times greater than that of erythrocytes. Certain biochemical parameters have beeneeXamined in an effort to find possible explanations for the variation of NTPH activity. In a search for cytoplasmic factors 14 which might alter NTPH activity, lysates of red cells from individuals with low, medium and high NTPH levels were mixed with partially purified red cell NTPH (5). The strictly additive nature of the results provided no evidence for the presence of cytoplasmic activators or inhibitors of NTPH. Likewise, the same report showed that the Km of NTPH for its substrate, ITP, is apparently not a factor in NTPH variability. The Km values which ranged from 2.1 - 5.7 X 10'5M did not reveal any pattern of differences for individuals with NTPH specific activities ranging from 12 - 124. Previous data from this laboratory (2) showed no evidence for NTPH electrophoretic variants in human red cells, rabbit red cells, and rabbit liver NTPH preparations. Harris and Hopkinson (15), in a screening for electrophoretic variants of NTPH, found only one such variant in a large number of red cell lysates from diverse racial origins. While they note observing quantitative differences in the amount of NTPH present in the screening study, no evidence for associated differences in the NTPH molecule itself could be found. The structural gene for NTPH is thought to be on chromo- some 20 in man, based on evidence of its synteny with adenosine deaminase (ADA) (previously determined to be on chromosome 20) in human-Chinese hamster somatic cell hybrids (16, 17). The pooled data of both reports reveal that 12 clones were Positive for both ADA and NTPH, 8 clones were negative for batlll’ 3 clones were positive for ADA and negative for NTPH 15 and 1 clone was negative for ADA and positive for NTPH. The locus for NTPH could not be excluded from chromosome 7 or 22 in either report so that, combined with some 6f the clone data (4 of 24 discordant), the assignment of the NTPH structural locus to chromosome 20 is somewhat unsure. In addition, data to be presented here regarding the subunit structure of NTPH suggest the existence of two NTPH structural gene loci. At this point there are no clinical conditions associated with low levels or total deficiency of NTPH. However, certain data have been published that may have some bearing on the effects of NTPH deficiency. Fraser gt_§l. (18), in a report from J. F. Henderson's laboratory prior to the collaboration with this laboratory, studied the accumulation of [14C] ITP in erythrocytes incubated with [14C] hypoxanthine. Erythrocytes from 5% of 80 normal individuals accumulated "high" amounts of [14C] ITP while 16% of 100 severely mentally retarded persons exhibited this characteristic. Although NTPH levels were not tested, the results of Soder gt 31. (5) suggest that the "high* ITP" individuals are probably low in NTPH activity. Furthermore, since the population survey of NTPH activity presented here shows that almost 18% of normal individuals have "low NTPH" specific activity, the "severely mentally retarded" data of Fraser gt a]: (18) 16 fit the population distribution of NTPH activity better than their "normal" data. Vanderheiden and Zarate-Moyano (19) examined NTPH levels in a psychiatric population. The psychiatric patients had significantly lower mean NTPH level than a normal population. Moreover, among paranoid schizophrenics, 5 of 78, or 6.4% had NTPH specific activity below 60 (umole Pi liberated per hr/g hemoglobin) while only 4 of 1063 (0.38%) normal individuals had such NTPH activity. These results must be questioned, however, because previous population data from Vanderheiden (I) showed that 4.1% of a normal sample group had NTPH activities below 60. Also, the assay conditions for NTPH, relying on endogenous in- organic pyrophosphatase for the production of Pi for a colorimetric assay, may lead to erroneous results. In an attempt to explain his observed differences between the normal and paranoid schizophrenic groups, Vanderheiden hypothesized that high levels of ITP present in NTPH deficient tissues inhibit the activity of glutamic acid decarboxylase (20). The latter enzyme has been found to be significantly reduced in activity in the brains of schizophrenics and psychotics (21). Further investigation is certainly required to establish the clinical effects of NTPH deficiency in humans. B. Genetic Mechanisms of Quantitative Variation of Enzyme Activity y 1. Biochemical bases of enzyme variation Variation in level of enzyme activity among individuals, such as that displayed by NTPH, is a common phenomenon. Much of this variation, especially where enzyme deficiencies are involved, is causative of health problems. These conditions range from those with mild effect, such as one form of gout, caused by partial deficiency of hypoxanthine- guanine phosphoribosyltransferase (22), to those such as Tay-Sachs disease, caused by deficiency of hexosaminidase A (23), which is generally lethal in infants before they reach three years of age. Stanbury EE.El° (24) list 147 human disorders in which a deficient activity of a specific enzyme has been demonstrated. In other cases of quantitative enzyme activity variation there may be no apparent deleterious effects of too much or too little enzyme activity. Catechol-0~methyltransferase, one of the two enzymes that catalyze the metabolism of the catecholamines, norepinephrine, epinephrine and dopamine, displays variation in activity in individuals over an eight-fold range, yet those individuals with the lowest activity exhibit no clinical signs of disorder (25). There are many factors that can affect the quantitative varniation of enzymes. Harris (26) suggests four categories 17 18 of methods by which a mutant gene can alter the level of enzyme activity in the cells of individuals: a) there may be structurally altered protein with defective or modified catalytic properties. The actual amount of enzyme protein present could be unaffected. b) there may be structurally altered protein whose catalytic activity is not significantly affected but whOse inherent stability is different than that of normal. The enzyme half-life may be reduced or increased. c) the rate of enzyme synthesis may be altered. This could involve mutation in the structural gene or in a possible regulatory gene. d) the mutation may affect the enzyme indirectly by affecting the intracellular concentration of activators or inhibitors. The quantitative variation of any one enzyme may also involve combinations of the above caused by one mutation. When the causes of enzyme activity variation ban be delineated, insights into the type of mutation responsible for the observed differences can be obtained. Most human variation seems to be the result of point mutations in the structural gene for the protein. In the simplest situation, a point mutation involves the substitution of one purine or pyrimidine base for another. This type of mutation is a? tha fifl “U ”a UV RVA V. a (J I 9‘ f ,- l 504. “O ‘- 3‘5 I' a... V I. “'5‘; fun l9 termed a missense mutation if the codon is changed such that it results in the replacement of one amino acid for another in the protein. The well-known substitution of valine for glutamic acid in hemoglobintSis an example of such a substitution (27). Missense mutations which cause variation in enzyme activity are generally positive for cross-reacting material (CRM). This refers to the method of preparing antibody to normal enzyme and the testing of the variant individual for the presence of protein that the antibody will react with. Protein is often made in normal amounts when missense mutations are involved, but that enzyme is not as physiologically effective. One or more physical parameters of the enzyme may be affected by the point mutation leading to the altered activity. Such parameters include increased or decreased binding affinity for or utilization of substrates and inhibitors, altered pH optima, altered electrophoretic mobility, and increased or decreased stability. Two other types of point mutations, nonsense and frameshift mutations can cause situations where little or no protein is detected as CRM. Nonsense mutations are the substitution of one base for another that results in a mRNA codon that is a stop signal rather than one that codes for an amino acid. Premature termination of synthesis of the protein molecule occurs; Chang and Kan (28, 29) recently identified a nonsense mutation in a 20 patient with homozygous 80 thalassemia. Although it was shown by cDNA . RNA hybridization that B-globin mRNA was present in the patient's reticulocytes, there was an absence of B-globin chain synthesis. A single nucleotide sub- stitution in the mRNA at a position corresponding to amino acid 17 resulted in an amber termination codon. Frameshift mutations, which involve additions or deletions of bases can also lead to premature termination of protein synthesis by producing stop signals in the altered reading of the codon. While nonsense and frameshift mutations in the structural gene produce CRM-negative results due to lack of protein production, other mutations in the structural gene can be CRM-negative in different ways. An altered protein can be produced that differs both in enzyme activity and its antigenicity, such that it will not react with antibody ‘to'the normal enzyme even though it may be present in Inormal amounts. Absence of CRM might result from an amino 3(11d substitution in the enzyme that leads to instability and rapid destruction of that enzyme. Alternatively, the decreased rate of synthesis of an enzyme, resulting in decreased CRM, could be the result of mutation in the structural gene. Such "structure-rate" relationships have been proposed by Itano and Boyer 3t _a_1_1_. (30 a 31) with regard to synthesis of variant hemoglobins. Among the possibilities is one that a point mutation might cilEIIISJera mRNA codon to that for a lees available tRNA u; ‘R nb‘ 38 1 C RAF VJ“ “F54 New 0 a 4 v Q a; ea 1‘. 21 species, thereby yielding a slower rate of translation of the enzyme. Polarity mutations in promoter or operator regions could cause deficient or decreased rates of trans- cription of the structural gene by altering various initiation and binding sites. An altered mRNA transcript of a mutated gene may not be properly processed in the nucleus and as a consequence does not reach the cytoplasm for translation. An intriguing possible cause for the lack of enzyme protein (CRM) in individuals with leSs than normal enzyme activity postulates that a mutation has occurred in a regulatory gene for that enzyme. Convincing evidence for a regulatory gene defect is not yet available in humans. Paigen (32) proposed the following criteria for identifying a regulatory gene mutation controlling enzyme levels: 1) the mutation must alter the rate of enzyme synthesis or breakdown 2) the mutation must be at a genetic locus separate from the structural gene for the enzyme ‘TC> date, no example unequivocally fulfilling both criteria has been detailed in humans. The genetic causes of variation of enzyme activity in individuals, as outlined above, can be illustrated well by tile; :numerous examples available in the literature of the ell-leidation of the biochemical bases of protein variation if! 11tnmans. Hemoglobin and glucoseé6—phosphate.dehydrogenase 22 (G-6-PD) are the non-enzyme and enzyme proteins, respectively, in which the greatest amount of variation has been found. MoKusick's catalog of human genetic variants (33) lists over 90 alpha globin and over 180 beta globin chain variants in hemoglobin for which the amino acid substitutions have been determined. Over 160 variants of G—6-PH had been reported. The following examples are provided to illustrate the general types of information that ban be ascertained in the study of genetic variation in enzyme activity. 2. Hypoxanthine-guanine phosphoribosyltransferase Hypoxanthine-guanine phosphoribosyltransferase (HGPRT) deficiency in humans results in two debilitating conditions -- gout and the Lesch-Nyhan syndrome (22, 34). Lesch—Nyhan syndrome is a lethal condition characterized by choreathe- tosis, spasticity, self-mutilation and mental and growth .retardation. Enzyme activity in erythrocytes of those patients with gout which results from partial deficiency cxf HGPRT is .01-17% of normal while that in Lesch-Nyhan syndrome patients is less than .004% of normal activity. (Ttluer cell types of these patients have similar levels of enzyme activity compared to normal. Immunological studies of Lesch-Nyhan syndrome patients by Rubin gt '31. (35) and Arnold 313 111;. (36) using antibody prepared against highly Purified HGPRT gave evidence that such patients had normal anl<=>unts of enzyme protein (CRM—positive) present despite tilee ‘rery low activity. This led to the conclusion that a and “3‘ VI. O 'w ‘H .d my race adv I EX: . l ‘.- bug -‘ ‘Rv “J" c '01 ‘V\ . .fi‘ 4.. 23 mutation in the structural gene was responsible for synthesis of a non-functional enzyme. However, Upchurch gt gt. (37) and Ghangas and Milman (38), studying some of the same patients as Arnold gt gt. (36), but using what they con- sidered to be more accurate procedures, found cross reactive material in only one of 16 Lesch-Nyhan patients. The methods utilized were based upon antibody competition experiments where measurement of cross-reacting material in HGPRT-deficient hemolysates involved inhibition of the immunoprecipitation of the normal enzyme. The lack of CRM indicates either a reduction in HGPRT protein concentration (by reduced rate of synthesis involving a mutation in a regulatory or structural gene, or by the production of a more labile HGPRT protein) or the presence of structurally defectiveeenzyme molecules with altered antigenic sites. LHowever, such conflicting evidence in CRM determination sniggests the need for care in interpretation of such data. Other evidence suggests that the CRM-negative HGPRT deficiencies in patients involve Structural gene mutations rather than regulatory gene mutations. The study of Arnold gt g_l_. (36) examined the it 11.19. loss of HGPRT activity during the ageing of erythrocytes. The fact that Protein synthesis does not occur in mature red cells of humans provides opportunity for enzyme variants with differences in stability to be expressed. The replenishment 24 of proteins in other tiesues masks such ig_yizg lability. Erythrocytes were fractionated by density (which is correlated well with age of the cells, the oldest cells being the most dense). While the erythrocytes from normal subjects lost little HGPRT activity in the course of ageing, old erythrocytes of Lesch-Nyhan syndrome patients had substantially less activity than the young cells. Cells of each of the Lesch-Nyhan syndrome patients were subsequently determined to be CRM-negative for HGPRT (37, 38). This altered enzyme half-life is consistent with a structural gene'mutation. The one patient that was determined to have a normal level of CRM (37, 38) was a "Km“ mutant. McDonald and Kelley (39) found that the HGPRT of this patient displayed abnormal sigmoidal kinetics in response to increasing con- centrations of the substrate magnesium 5-phosphoribosyl-1— perphosphate (MgPRPP) in contrast to the hyperbolic Icinetics observed with the normal enzyme. Normal Michaelis- Menten kinetics were observed using hypoxanthine or guanine as the variable substrate, but the apparent Km's for MgPRPP, hypoxanthine and guanine were all an order of magnitude higher for this Lesch-Nyhan HGPRT than for the normal snazzgfme. 25 The heterogeneity in HGPRT-deficiency includes patients with enzyme of altered thermostability, such as patients with gout studied by Kelley gt gt. (40). Some mutant enzymes examined were inactivated much more rapidly than normal when heated at 80°C. However, the HGPRT of two patients from the same family was more heat resistant than the normal enzyme. Also, in the patients of this family hypoxanthine was utilized at a rate of 19% of normal, while guanine was utilized at a rate of 0.5% of normal. Most HGPRT utilizes both equally. Once again, a mutation in the structural gene is probably responsible for these characteristics. 3. Glucose-6-phosphate dehydrogenase Glucose-6-phosphate dehydrogenase, (G-6—PD), as previously mentioned, exhibits the most variability thus far cdetermined of any enzyme protein. Deficiency of G-6-PD (Lenses drug-induced hemolytic anemia (41), the severity of vflhich is directly related to the extent of the activity deficientyc. Beutler (42) lists the reported variants up t1: ‘the year 1978 along with their biochemical characteristics. These variants include those that differ from normal G-6-PD in electrophoretic mobility, Km's for G-6—PD and NADP, utilization of other substrates (2-deoxy G—6—PD and deamino “UKIDI’I, K1 for the inhibitor NADPH, thermostability, and pH optimum. Some variants such as the A— and the Mediterranean d r A" .I. 26 types of G-6-PD are quite common, with gene frequencies, respectively, of .11 in American blacks and .2 - .4 in certain parts of Greece. Since G-6-PD is X-linked, these are also the frequencies of males which express the enzyme deficiencies. The A— type G-6-PD has an activity which is 8 - 20% of normal, while the Mediterranean type has an activity which is lees than 7% normal (33). The A- type of G-6-PD also exhibits an altered electrophOretic mobility while the Mediterranean type differs from normal G-6-PD in that it has lower Km's for G-6-PDand NADP, utilizes 2-deoxy G-6-PD and.deamino NADP at greater rates and is more heat labile (42). Piomelli gt_gt. (43) examined the rate of decline of enzyme activity ig_!izg during the maturation of red cells via fractionation of red cells by density (and hence by age). They found that the youngest fractions from A- erythrocytes aontained G-6-PD activity nearly equal to that in normal cezlls. However, the activity in A- cells dropped dramatically with age such that the enzyme half—lives were estimated to be 13 days for the A- enzyme compared to 62 days for the ncxrjnal G-6-PD. Their data for the Mediterranean type was rust: as clear because of the low activity, but nonetheless it: teas suggested that the synthesis rates for the A- and Mediterranean G—6-PD variants were normal and that the enZYIne deficiency resulted from it vivo inactivation. G“) (I) 0‘ 4) I.” 'I l (I) 27 Studies by Morelli gt gt. (44) of the Mediterranean type of G-6-PD measured G-6-PD protein levels by radioimmunoassay and enzyme activities of G-6-PD of red cell preparations separated according to density (age). Their results con- firmed those of Piomelli gt gt. (43) with regard to tg_zttg inactivation of the variant G-6-PD. The Mediterranean type of G-6-PD variant was found to have approximately normal amounts of enzyme protein in the youngest cells, but this G-6-PD variant was less catalytically active than normal G-6-PD —- a phenomenon which compounded the activity loss during the red cell ageing. The Hektoen variant of G-6—PD appears to be unique among human enzymes in that the enzyme activity is increased to 4 - 5 times that in the normal person (45). Carrier females have activity levels ranging from l-l/2 to 4 times .normal. The increased activity of G-6-PD was also shown to be present in the leukocytes of these individuals. Of all biochemical parameters tested, only a slight change in electrophOretic mobility was noted (46). All other properties <1f"the enzyme seemed to be identical to the normal enzyme. This included a normal specific activity as measured by quasirititative immunologic neutralization. Yoshida (47) Purified normal G-6-PD and the Hektoen variant. Peptide mapping and subsequent amino acid analysis of the differing Spots determined that the Hektoen variant was the result .V 6'. av) babe hr 7'!- 1. v . ‘- 3:. u.‘ ‘N 'I ID I n 'L-l 2: i .k‘ ': 28 of a substitution of a tyrosine residue for a histidine residue. This substitution was consistent with the altered electrophoretic mobility. It is likely, then, that a structural mutation has in some way affected the rate of transcription and/or translation of this particular variant G-6-PD allowing increased synthesis of enzyme. 4. Pyruvate kinase Like G-6-PD, pyruvate kinase is an enzyme which exhibits considerable heterogeneity. Deficiency of pyruvate kinase causes hemolytic anemia (48). Such deficiency appears to be inherited as an autosomal recessive trait with homo- zygotes showing 5 to 25 percent of normal activity. Heterozygotes have about half the normal pyruvate kinase activity (49). Black gt gt. (50) studied eight patients representing five families with pyruvate kinase deficiency— associated hemolytic anemia. Each of the five families Exresented unique combinations of differences from the normal enzyme in activity levels, electrophOretic mobility, inununologically detectable protein, Km's for the substrates phosphoenolpyruvate and ADP, effect of the positive allosteric effector fructose-l, 6-diphosphate, and thermostability. An intereSting mutant pyruvate kinase was studied by Adachi gt gt. (51) . The enzyme pyruvate kinase activity of the Patient's 'hemolysate was only 5 - 7% of normal. The Km for phosphOen'olpyruvate (PEP) was decreased and there ‘fl31763 Eiltered allosteric properties with PEP compared to “I. A u c “J 29 the normal enzyme. While normal pyruvate kinase is a tetramer, gel chromatography of the mutant enzyme reVealed the mutant pyruvate kinase was easily dissociated into monomers.‘ The monomers possessed a low but measurable enzyme activity. Normal pyruvate kinase monomers obtained by sonic disruption also have low activity. The authors suggested that the low enzymic activity found in the mutant tetramers compared to normal tetramers was a result of a lack of normal conformational change in the mutant enzyme upon aggregation of the subunits. 5. Red cell acid phosphatase‘ Electrophoresis has undoubtedly been the methOd most utilized for searching for mutant forms of enzymes and proteins. However, since only about one third of all base substitutions will result in altered electrophoretic mobility (52), much.variation remains undetected using this analytical tool alone. While electrophoretic variationiis common, many alleles detected in this manner give no evidence of quantitative variation in enzyme activity. Red cell acid Phosphatase is an early classic example of a common electro- Phoretic polymorphism with associated quantitative differences in enzyme activity between the alleles (53) . Three electro- FHJ£>Jretic alleles A, B, and C render Ggenotypes which have the following frequencies: 9.6% type A, 48.3% type BA, 34 - 3% type ’B, 2.8% type CA, 5.0%. type 'CB and approximately 30 0.2% type C. Analysis of the distributions of activity of red cell acid phosphatase in hemolysates of individuals with different phenotypes reveals that each of the alleles has a different activity level and that the activity appears to be additive in heterozygotes. Thus, the A phenotype with a mean activity of 122.4 (umoles disodium p—nitrophenyl phosphate cleaved at 37°C/30 minutes/ mg hemoglobin) and the B phenotype with a mean activity of 188.3 predict an activity of 155.35 for the BA phenotype (1/2 A + 1/2 B) which matches well the observed mean of 153.9. The C allele contributes the highest activity. The overall activity distribution, however, is essentially continuous and unimodal. Interestingly, thermostability studies indicate that the C allele is also the:most stable followed by the B allele, and then the A allele (54). 6. Peptidase A Lewis (55) made use of electrophoresis in a somewhat different manner in the examination of variation of red czeell peptidase A. Though seven eleCtrophoretic alleles were Icrlown in the red cell, none segregated in families with low £>€azrtidase A activity. However, in white blood cells where title! activity was not depressed, an electrophoretic allele Was discovered that was associated perfectly with the low activity pattern of inheritance in red cells. This indicated that a structural, gene variant was responsible for the Sltléalutitative variation. 31 7. Catechol-O-methyltransferase One enzyme that exhibits variation in activity quite similar to that of nucleoside triphosphate pyrophospho- hydrolase is catechol-O—methyltransferase (25). As mentioned previously, there are no clinical manifestations resulting from low activity of this enzyme. A population distribution revealed an apparent bimodal distribution of activity with 22.9% of the total population in a "low" activity mode. Activity in the low region covered a 2—fold range while that in the high.region covered a 3-fold range. Family studies were conducted which indicated that low activity was inherited as an autosomal recessive trait. Low activity individuals were therefore homozygous, while the high activity range was comprised of both heterozygotes and homozygotes for the high activity allele. Subsequent studies of catechol-o-methyltransferase from.low activity individuals found no evidence of endogenous enzyme inhibitors or ar:tivators, differences in Km's for the substrates dehydro- xybenozoic acid and S-adenosyl-l-methionine, or differences -111. concentrations of inhibitors at which 50% inhibition o<=<=urred (56). However, examination 95 thermostability reT-Iealed that the enzyme in lysates of low enzyme activity ‘VEIEB :more heat labile than that in the lysates with normal leixreels of activity. Heterozygotes, though, could not be distinguished from homozygotes of the high activity range :v . .9 a :w 'Iha 1. o v m D» u u e o e)“ :3 av .. s :1. .hu 5» nu .u . Q 1‘ a s a I a an. vNU ,0 n DU oqlo 32 by the thermostability criterion, since the rates of thermal inactivation are not sufficiently different. 8. Galactokinase Some population distributions of enzyme activity like that of catechol-o-methyltransferase where a bimodal activity distribution is observed, can be explained by a simple two allele system of inhertance. Other distributions of enzyme activity have proven to be much more complex. The common polymorphism of red cell phosphatase already mentioned presents a unimodal distribution of activity, but the ability to separate the three alleles involved by their electrophoretic mobility allows ready interpretation of the pattern of inheritance. The variation in activity of the enzyme galactokinase (GALK) in Caucasian and black populations provides an example of genetic analysis of unimodal distributions of activity in the absence of (qualitative biochemical differences. A population survey of GALK activity in whites and blacks showed highly Significant differences between the mean red blood cell GALK activities of the two groups; the mean for blacks being about two thirds that of the white population (67) . Both distributions were unimodal. While the data on whites $11<>~wed adequate fit to the normal distribution, that for I’luélczks showed significant departure, being skewed toward hiQ’l‘ier GALK activities. Two common alleles were postulated, GALKA and GALKP. GALKA was thought to be the normal 33 allele of white populations, such that the great majority were homozygous. GALKP was thought to be present in the black population, which in combination with GALKA, produced three genotypes: GALKA GALKA, GALKA GALKP, and GALKP GALKP. Spielman gt gt, (58) used various statistical methods including chi-square analysis and maximum likelihood estimations to fit genotype distributions to the over- all population distribution of GALK activity in blacks. The GALKA allele was estimated to be present in the black population at a frequency of .217. The proposed GALK genotypic activity ranges obtained from the statistical analyses mentioned above suggest where in the distribution of activities one might expect to find samples of individuals nearly homogeneous in genotype for study of possible bio- chemical differences of the gene products GALKA and GALKP. III. Reagents, Materials and Methods A. Reagents l. Commerical The following chemicals were obtained from Sigma Chemical Company, St. Louis, Missouri: dithiothreitol, N-ethylmaleimide, p-hydroxymercuribenzoate, yeast inorganic pyrophosphatase, bovine serum albumin, rabbit muscle pyruvate kinase, bovine pancreas chymotrypsinogen A, calf thymus deoxyribonucleic acid, acrylamide, N,N'—methylene— bis-acrylamide, N,N,N', N' tetra methylethylenediamine, Sephadex 675-40, and CM Sephadex C—50-120. The sodium salts of the hypoxanthine, guanine, xanthine and adenine ribonucleoside—S'—triphosphates and deoxyinosine 5' triphosphate were obtained from P—L Biochemicals, Milwaukee, Wisconsin. Sodium dodecyl sulfate and acrylamide were obtained from Pierce Chemical Company, Rockford, Illinois. Acrylamide was also obtained from Miles Laboratories, Elkhart, Indiana, as was ammonium persulfate. Yeast inorganic pyrophosphatase was also supplied by Nutritional Biochemicals and U.S. Biochemical Corporation, both of Cleveland, Ohio. 34 dea: .A " Vol 1 3*»: . 50‘ 35 Bovine heart lactate dehydrogenase and bovine pancreatic deoxyribonuclease were purchased from Worthington Biochemical Corporation, Freehold, New Jersey. Isoelectric focusing ampholytes were purchased from the following sources: Ampholine 3-6 was obtained from LKB-Produkter AB, Bramma, Sweden, Biolyte 4—6 from Bio-Rad Laboratories, Richmond, California, and Pharmalyte 4-6.5 as well as Sephadex G-100 (40-120) were obtained from Pharmacia Fine Chemicals, Uppsala, Sweden. Miscellaneous reagents were obtained from the following sources: 2-mercaptoethanol and Photoflo 200 solution were purchased from Eastman Kodak Company, Rochester, New York. Iodoacetamide was obtained from Aldrich Chemical Company, Milwaukee, Wisconsin. DE 52 anion exchange resin was supplied by Whatman Biochemicals, Maidstone, Kent, England. Coomassie Brilliant Blue R.250 was obtained from Bolab Laboratories, Glenwood, Illinois and Xylene Brilliant Cuyanin G was obtained from ICN Pharmaceuticals, Plainview, New York. Hycel cyanmethemoglobin reagent (No. 116E) was <>k>tained from.Hycel, Inc., Houston, Texas. Bis-acrylylcystamine, sl’nthesized by the method of Hansen (59) was the generous gi ft of H.P. Hershey of this laboratory. All other chemicals used were reagent grade; 36 2. Recrystallization of acrylamide and N,N'-methylene- bis-acrylamide Acrylamide and N,N'-methylene-bis-acrylamide obtained from Sigma Chemical Company were recrystallized in the following manner: 100 g of acrylamide was dissolved in 200 ml acetone at 40 - 50°C. 'A small amount of Celite was added, then the mixture was filtered through Whatman #1 filter paper in a Buchner funnel with vacuum. The filtrate was placed at 4°C overnight. The crystals were filtered off in a Buchner funnel and placed in a dessicating jar under vacuum to thoroughly dry. 15 g of N-N'-methylene-bis-acrylamide was dissolved in 800 ml acetone at 40 - 50°C. The remaining procedure was as above. B. Materials 1. Commerical Vacutainers and needles were obtained from Becton- Dickinson, Rutherford, New Jersey. Heparinized micro- hematocrit capillaries and 0.45 micron filters were Purchased from Arthur H. Thomas Company, Philadelphia, Pennsylvania. PM 10 ultrafiltration membranes were PUrchased from Amicon, Lexington, Massachusetts. Cellulose nitrate centrifuge tubes 5/8 inch diameter 37 x 2-1/2 inches and 1/2 inch diameter x 2 inches were obtained from Beckman Instruments, Inc., Palo Alto, California. Whatman filter papers were obtained from Whatman, Inc., Clifton, New Jersey. 2. Special preparation of glass tubes for poly- acrylamide gel electrophoresis Glass gel tubes 5 mm i.d. X 125 mm were cleaned in Chromerge solution, rinsed extensively with tap water and finally with distilled water and dried. The tubes were then placed in a 0.5% Photoflow 200 solution overnight and dried. 3. Biological-~Blood samples a. Population survey Random blood samples from a Caucasian population were obtained from the American Red Cross, Lansing, Michigan. b. Families selected for genetic analysis All families presented in this study, except for those provided through J .F. Henderson, were chosen at random for their NTBH phenotype, termed "complete Selection" by Morton (60). All were Caucasian whose ‘afiraiilability‘was the only criterion for selection. Blood salIlflples were drawn by venipuncture with the aid of Va<3'utainers containing sodium heparin. NTPH analyses were perfomed within 3 days. Remaining red cells were stored 1’1 11m1quid nitrogen for later use. Written informed consent '7 . :92. n . “‘1 Va. 38 was obtained from all participants. This study was approved by the Michigan State University Committee on Research Involving Human Subjects, September 12, 1977. c. Other families obtained Blood samples from 15 families from Dr. J.F. Henderson's laboratory, University of Alberta Cancer Research Unit, Edmonton, Alberta, Canada, were shipped as packed cells in sealed ampules on dry ice. The propositi for these families were individuals whose red cells synthesized high levels of [14C2 ITP from [14C] hypoxanthine. These families, then, were not randomly ascertained and will be noted as such in the results and discussion. Blood from Family 00 was obtained through Hurley Hospital in Flint, Michigan. Blood from Families NN and PP was obtained at Michigan State University. d. Psychiatric patient samples Blood samples of normal personnel and of pgychiatric patients with varying diagnoses were obtained from Dr. R. Kobes, Department of Clinical Psychopharmacology, National Ilistitutes of Health, St. Elizabeth Hospital, Washington, D.C. e. Miscellaneous samples used for study Unites of outdated whole blood were obtained from the American Red Cross, Lansing, Michigan. Blood samples from Michigan State University staff and students were obtained as needed. 39 C. Methods 1. Analytical a. NTPH, phosphate, hemoglobin, and protein analyses NTPH catalyzes the general reattion: NTP + H20 + NMP + PPi. The pyrophosphate produced during the incubation of NTPH with ITP was hydrolyzed by yeast inorganic pyrophosphatase according to the method of Chern gt gt. (1). The inorganic phosphate thus formed was analyzed colori- metrically by the method of Rathbun and Betlach using K2HPO4 as a reference standard (61). Unless otherwise noted, the assay mixtures contained 50 mM alanine buffer (pH 9.5), 10 mM MgClz, 1 or 2 mM DTT, 1 unit of yeast inorganic pyrophosphatase, 0.5 mM ITP and the NTPH-containing solution. The use of 2 mM DTT enhanced the reproducibility of the determinations, especially when (a great number of assays were donducted simultaneously. Chantrol mixtures were incubated concurrently to account for tflie traee phosphate contamination in both the ITP and NTPH- containing solutions. ATP at a concentration of 0.5 mM Vfiifii sometimes substituted for ITP to test for the presence of nonspecific phosphates. In analysis of NTPH hydrolysis 0f other triphosphates the concentration used was also 0. 5 mM. All analyses were conducted in triplicate. 40 Following the incubation of these mixes for 20 minutes at 37°C, the reactions were terminated by the addition 6f 2.2 ml cold 7.27% trichloroacetic acid. The precipitated protein was removed by centrifugation (1000 g for 5 minutes at 4°C) and the supernatants were decanted into phosphate-free test tubes for inorganic phosphate determination. Phosphate analysis was accomplished by adjusting the pH to ca. 4.5 with 1.88 ml 3M acetate buffer (1 : 1 mixture of 3.0 M sodium acetate and 3.0 M acetic acid) to which 1/10 part 37% formaldehyde had been added to diminish the interference of sulfhydryl reagents in the color development stage. ‘Then 0.2 ml 2% ammonium molybdate followed by 0.4 ml 6.75 mM stannous chloride were added to each tube. After 15 minutes absorbance at 700 nm of each of the solutions was determined with a Gilford 300 microsample spectrophotometer. A unit of NTPH activity is defined as that amount of enzyme which hydrolyzes one nmole of ITP at 37°C in the standard 20 minute incubation. The specific activity may be expressed as units per mg protein, units per mg hemoglobin, or units per cell as indicated. The hemoglobin concentration in hemolsates was determined by the method of Austin and Drabkin (62) . The de‘-’-e3|’:'tuinations were faciliated by the use of Hycel cyan- methluoglobin reagent. The absorbance at 540 nm of a 1 mg/ml cYerunethemoglobin solution is 0.718. 2'19 graz' an Vii \ flea} any. 1 vv ‘T 41 Other protein analyses were carried out by the method of Lowry gtht. (63) with bovine serum albumin as the reference standard. 5. Sucrose density gradient centrifugation The sedimentation behavior of human red cell NTPH was examined in sucrose density gradients essentially by the method of Martin and Ames (64). Linear 5 to 20% sucrose gradients were prepared in 0.05 M Tris-HCl buffer containing mM glutathione at pH 7.5 at 4°C. Each chamber of the gradient mixer contained 2.3 ml of the respective 5% and 20% sucrose solutions. The gradients were prepared in Beckman 122 inch diameter X 2 inch cellulose nitrate centrifuge tubes. The 0.1 m1 sample loaded on top of the gradient contained the following: 85 pl of partially purified NTPH $13,000 units/mg) purified through the CM-Sephadex C-50-l20 step as described by Morris (8), 5 ul of human hemoglobin solution (29 mg/ml, prepared as a hemolysate in H30) and 10 ul of a beef pancreas deoxyribonuclease solution (1mmg/O.l ml, 1200 units/mg). The centrifugation was carried out using a: Beckman L2--65 ultracentrifuge with a SW 50.1 rotor, at 5C),000 RPM for 16 hours at 4°C. Fractions of 10 drops were cc>llected using a Gilson Microfractionator. NTPH activity was determined as previously described. (See Section III - C. 1. a.) . The hemoglobin containing fractions were Hbrzjitored spectrophotometrically by absorbance at 415 nm. 42 Deoxyribonuclease activity was assayed by the method of Kunitz (65) based upon the increased absorption at 260 nm observed during the depolymerization of DNA by the enzyme. Calculations of S value for NTPH were based on the S values for human hemoglobin and deoxyribonuclease as determined by Chiancone gt gt. (66) and Lindberg (67), respectively. c. Isoelectric focusing in sucrose density gradients The apparatus used for isoelectric focusing in sucrose density gradients consisted of a 10 ml burette with the tip cut off, a 5 cm diameter glass tube attached to the top to serve as the cathode reservoir, and a capillary tube con- taining a polypacrylamide plug attached to the bottom of the burette. This apparatus was placed inside a 2 liter graduated cylinder with a stir bar which served as the anode reservoir. The polyacrylamide plug was prepared as per the separating gels described in Section III. C. l. e. A dense electrolyte solution placed in the bottom of the burette was comprised of 1.5 mg GSH, 0.2 ml IM H2504, 3 9 sucrose and 3 ml H20. The gradient (5—50% sucrose) which was Exaured on top of the above dense solution consisted of l..7 mg GSH, 0.14 ml Pharmalyte 4-6.5, 2.7 9 sucrose and 3..4.m1 H20 in the mixing chamber of the gradient mixer armd.l.7 mg GSH, 0.14 ml Pharmalyte 4-6.5, 0.27 9 sucrose, and 5.1 m1 H20 in the other chamber. The sample to be Jh>éaded in the gradient was placed in the light gradient solution replacing an equal volume of water. The gradient 43 was topped by the cathode solution consisting of 36.8 mg GSH, 6 ml 1M NaOH and 114 ml H20. The anode solution which filled the graduated cylinder consisted of 88 ml 1M H2804 and 2112 ml H20. The isoelectric focusing was run at 300 V for 30 minutes, then increased to 500 V for 24-48 hours. 0.2 ml fractions were collected using a Gilson Microfractionator and were analyzed for NTPH activity, absorbance at 280 nm, and pH measured at 4°C without dilution. d. Isoelectric focusing in polyacrylamide gels Isoelectric focusing in polyacrylamide gels was done by modification of a procedure suggested by Bio-Rad Laboratories. Gels of 7.5% with 2.5% crosslink were prepared in the following manner: In 25 ml total volume were added 6.25 ml monomer solution (29.2% acrylamide, 0.75% Bis), 1.2 ml 50% glycerol, 1.25 ml ampholyte (either Pharmalyte 4-6.5 of Biolyte 4-6, final concentration 2%). and 0.5 ml 0.02% riboflavin monophosphate solution. Gels were poured in glass tubes 0.5 x 12.5 cm to 11 cm in length, crverlayered with.water and allowed to polymerize overnight uJader fluorescent light. The upper reservoir solution was 0..06 N H2804 and the lower reservoir solution was 0.6 N NaOH. Total samplessizes of up to 400 pl were loaded on tfll£e gels as 25% sucrose solutions with ampholyte at con- cen tration of 0.4%. The samples were overlayered with 105; sucrose also containing ampholyte at 0.4% concentration. 44 The anode was placed in the upper reservoir and the samples were run at constant voltage of 20 V/cm for 18 hours. Hemoglobin migrated the length of the gel to the cathodic end while NTPH focused nearer the anode. The gels were extracted from the tubes and stained for activity and protein as described below (See Section III. C. 1. h., and III. c. 1. i.) e. Polyacrylamide gel electrophoresis Analytical polyacrylamide gel electrophoresis was carried out using modifications of the method of Gabriel (68). Best sebolution of the proteins that migrate near NTPH was achieved using 7.5% separating gels with 2.5% of total monomer comprised of N, N'-methylene-bis-acrylamide (Bis). The separating_ge1 formulation consists of 1 part separating gel buffer (1.5 M Tris-HCl containing .12% N, N, N', N'- tetramethylethylenediamine (TEMED) pH 8.9 at 25°C), 1 part monomer solution (30% acrylamide, .736% Bis), and 2 parts catalyst solution (0.14% ammonium persulfate). The solution was deaerated briefly and separating gels were ENDnred to 9 cm in 0.5 X 12.5 cm glass gel tubes. Stacking gels of 1 cm were added following polymerization of the separating gel. The stacking, gel formulation consists of 5 parts stacking gel buffer (0.25 M Tris-HCl containing o468 TEMED pH 6.7 at 25%C) , 5 parts monomer solution u; be- )1? “s 45 (20% acrylamide, 5% bis), 9 parts water, 1 part riboflavin monophosphate solution (0.02% in H20), and 20 parts 40% sucrose solution. The stacking solution was poured on the separating gel, layered with H20, and polymerized with fluorescent light. The electrophoresis buffer is .025 M Tris-glycine pH 8.3 at 25°C. A sulfhydryl reagent such as 2-mercaptoethanol at mM concentration in the gel formulations and buffer was routinely included for maintenance of NTPH activity for activity staining. The use of 2- mercaptoethanol appeared to have little effect on gel polymerization or protein separation. Pre-electrophoresis of the separating gels was found to have no positive effect on NTPH activity in the gels, so that most gels were run without such a procedure. Samples up to 150 pl were applied to the gels beneath the buffer in 10% glycerol with 3-6 ul of 0.01% bromphenol blue used as tracking dye. Electrophoresis was carried out at 4°C with a constant current of 2.5 mA applied for each gel. The length of electrophoresis was generally 4—5 hours. Staining for Errotein and NTPH activity was carried out as described below (See Sections III C. l. h. and III. C. l. i.). f. Sodium Dodecyl Sulfate-polyacrylamide gel electrophoresis For analysis of molecular weights, sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis was i”: Vdi if: ‘1. "A N‘H ‘ 46 carried out by the method of Weber gtfgt. (69). Gel buffer was prepared by titration of 0.2M NaZHPO4 by 0.2M NaH2P04 to pH 7.2 at 25°C. Each solution contained 0.2% SDS. The formulation of the 10% acrylamide gels preparation was 4.5 ml acrylamide solution (22.2% acrylamide, 0.6% Bis), 5.0 ml gel buffer, 0.5 ml 1.5% ammonium persulfate, and 15 ul TEMED. The gel solution was deaerated for about one minute before addition of the ammonium persulfate and TEMED. Gels were poured to 9 cm in length and layered with water. Polymerization occurred within 20 minutes. A sample buffer was prepared comprised of 1 ml gel buffer, 1 ml 20% SDS, 200 ul 2-mercaptoethanol, and 2.8 m1 H20. Protein samples were mixed three parts to one part with sample buffer and heated at 100°C for one minute. Heated preparations, 50—150 ul, were then mixed with 5 pl .01% bromphenol blue, one drOp glycerol, and 1/10th part 2~mercaptoethanol and loaded on the SDS gels. Electrophoresis buffer was made of equal parts gel buffer and water. The electrophoresis was conducted at room temperature with BmA constant current applied peer gel, generally for 4-1/2 - 5 hours. The gels were eactruded from the tubes, rinsed briefly and stained as described below (See Section III. C. 1. h.). Molecular weight standards used and their subunit weights were the following: bovine serum albumin-~68,000 rabbit '— 47 muscle pyruvate kinase--57,000, bovine heart lactate dehydrogenase-~35,000, bovine pancreatic deoxyribonuclease-- 31,000, bovine pancreas chrmotrypsinogen A--25,000 and human hemoglobin--15,500. Generally, 1.5 ug of each marker was loaded on the sample gel. NTPH subunit molecular weight was estimated graphically from a plot of the relative mobility of the standards versus the log of their subunit molecular weight. g. Solubilization of polyacrylamide gels To facilitate subunit molecular weight analysis of NTPH, solubilizable gels using Bis-acrylylcystamine (BAC) rather than Bis—acrylamide for cross-linking purposes were utilized. These gels can be dissolved using 2-mercaptoethanol or dithiothreitol. Bis-acrylylcystamine was synthesized by the methOd of Hansen (59). Following the suggestions of Bio-Rad Laboratories Product Information Sheet 2045, the formulation for 7.5% BAC gels is as follows: 5 ml stock acrylamide-BAC solution (13.5% total acrylamide/5% crosslink, 6.41 g acrylamide, 2.5 g BAC, H20 to 50 m1) , 2.25 ml gel btuEfer (1.5 M Tris-HCl pH 8.9 at 25°C), 180 pl TEMED and l- 6 ml ammonium persulfate (0.1125% solution). The gels are pre-electrophOresed overnight at 100V using .375 M Tris-HCl with mM GSH pH 8.9 at 25°C.as the pre-electrophoresis b11f fer. The stacking gel and electrophoresis buffer are the same as those used for the regular polyacrylamide gels 48 with the exception that GSH is substituted for 2-mercapto- ethanol in the buffer. Staining for activity and protein are carried out as with the Bis-acrylamide-containing gels (See Sections III. C. l. h. and III. C. l. i.). The visualized protein bands are cut from the gels and solubilized with 90 ii 1.0 M DTT. At this point, the solubilized gel slices are quite viscous. They are treated as other samples in preparation for SDS-polyacrylamide gel electrophoresis (See Section III. C. l. f.). However, the viscosity remains such that a more uniform application of the sample to the SDS_gel is facilitated by the use of a 1 m1 disposable syringe equipped with a 20 gauge needle. Alternatively, protein was recovered from Bis-acrylamide crosslinked polyacrylamide gels as follows: the stained protein bands were cut from the gels and the gel slices homogenized in 5 ml 1 N ammonium hydroxide, for 30 minutes. The polyacrylamide debris was pelleted by centrifugation at 4000 g for 10 minutes. The supernatant solution was lyophilized, washed once with water, and lyophilized again. The lyophilized protein samples were resuspended in SDS gel buffer in preparation for SDS-polyacrylamide gel electrophoresis as described in Section III. C. l. f. 49 h. Protein staining of polyacrylamide gels The protein staining technique used for native polyacrylamide gel electrophoresis and isoelectric focusing was that of Blakesley and Boezi (70). This method utilizes Coomassie Brilliant Blue G250 (xylene brilliant cyanin G), is rapid, and requires little destaining. Gels were placed directly in the staining solution overnight and then stored in water. The gels were scanned densito- metrically at 600 nm using a Gilford 24008 spectrophotometer equipped with a linear transport system. For sodium dodecyl sulfate (SDS)—polyacrylamide gels, a different stain was used, as that above caused precipitation of sodium dodecyl sulfate, resulting in an opaque background. The staining method of Weber gt gt. (69) was used for SDS gels. .This stain utilizes Coomassie Brilliant Blue R 250 in methanol and acetic acid. SDS gels were rinsed briefly in water, then placed in the staining solution for l-1/2- 2 hours. Destaining was accomplished in a solution of 5% methanol and 7.5% acetic acid with the aid of a Buchler shaker. Gels are stored in 7.5% acetic acid. i. Staining of polyacrylamide gels for NTPH activity NTPH is localized on polyacrylamide gels by the following method: after electrophoresis or isoelectric focusing, the gels are placed in a tube containing 20 ml of an NTPH reaction mix comprised of 50 mM B-alanine buffer 50 pH 9.5 at 37°C, 10 mM MgC12, 2 mM DTT, and 0.5 mM ITP. The gels are incubated 1-2 hours at 37°C in this solution, then transferred to a staining solution modified from Harris and Hopkinson (15). The staining solution consists of 2.5% ammonium molybdate and 5% ascorbic acid in 4N H2504. NTPH activity is visualized as a dark blue staining area on a lighter blue background. The intensity of the staining is proportional to the amount of NTPH loaded on the gel. No staining is seen using ATP in place of ITP as the substrate. j. Inhibition studies with N-ethylmaleimide, iodoacetamide, and p-hydroxymercuribenzoate Hemolysates (See Section III. C. 2. b.) used in the inhibition experiments were dialyzed in 3 ml aliquots 15-17 hours against 2 - 1 liter portions of 25 mM Tris-HCl buffer containing mM 2-mercaptoethanol, pH 7.4 at 4°C. Purified NTPH (See Section III. C. 2. b.) was diluted in 24 volumes of the above buffer and used directly. To 400 pl lysate or NTPH preparation were added 100 pl thiol reagent and H20 in combination to produce the desired inhibitor concentrations. These mixtures were incubated 30 minutes at 4°C. A 50 pl aliquot of the incflbation was placed in 1‘ml of reaction mixture for analysis of NTPH activity (See Section III. C. l. a.). 51 k. Red cell separation by age Fractionation of red blood cells by their density (age) was accomplished by the method of Murphy (71) as modified by Cohen gtht. (72). Twenty milliliters of blood are drawn from donors using Vacutainers with sodium heparin as anticoagulant. The blood, used fresh, was centrifuged at 3500 rpm for 20 minutes at 4°C using a Sorvall Model RC2 centrifuge with an 88-34 rotor. The plasma was removed and saved, while the buffy coat was aspirated with as few red cells as possible. The red cells were resuspended in the plasma and centrifuged as before. The plasma and remaining buffy coat were removed by aspiration and dis- carded. Hematocrits determined at this point were 89-92%. The red cells were transferred to a 1/2 x 2 inch cellulose nitrate centrifuge tube and centrifuged at 25,400 rpm (39,000 g) for one hour at approximately 30°C in a Beckman L2-65 ultra—centrifuge using a Type 50 rotor. (The rotor had been warmed to 30°C beforehand-~a step necessary for proper separation of the red cellsl) The remaining plasma was aspirated carefully and discarded. Usually 7 - 7.5 ml of packed cells remained at this point. Six fractions were obtained using a Beckman tube slicer-- the top and bottom 10% of red cells and 20% fractions in Jbetween. The fractions were lysed in nine volumes of mM 52 DTT and were analyzed for NTPH activity and hemoglobin con- centration as previously described (See Section III. C. l. a.). 1. Thermostability experiments All thermostability experiments utilized hemolysates of red cellsiin nine volumes of 2 mM DTT which resulted in hemoglobin concentrations of 29-32 mg per ml of hemolysate. At time 0, aliquots of a subject's hemolysate were placed in thick-walled glass ignition tubes (which had been pre- incubated to achieve proper temperature) and incubated for the designated time at 65°C in a high capacity Blue M Magni-Whirl MW-llSZSSA water bath. The tubes were then placed directly in an ice and water bath to cool and centrifuged at 1000g for five minutes at 4°C. The super- natant solutions were collected and analyzed for NTPH activity as previously described (See Section III. C. l. a.). .m. Computer analysis of data Lines were fitted to data points for the thermo- stability studies by a least squares computer program stored in the CDC 6500 which is available through a teletyps intercom system at Michigan State University. 2. General Procedures a. Preparation of hemolysates For analysis of NTPH activity, whole blood was czentrifuged at 4000g for five minutes at 4°C to sediment the cells. The plasma and buffy coat were removed by 53 aspiration, and the red cells washed once or twice with 3-5 volumes 0.9% NaCl and collected by centrifugation as before. The washings were discarded and the packed cells lysed with nine volumes cold 1 or 2 mM DTT. The cell debris was removed by centrifugation at 30,000 g for 15 minutes and the supernatant was used for NTPH and hemo- globin analyses. In some cases the hemolysate was prepared more concentrated or dilute than described here. b. Purification of human erythrocyte NTPH NTPH from outdated human blood obtained from the Red Cross was purified through the CM-Sephadex C50-120 chromatography step has described by Morris (8). The C50 eluate was dialyzed against 2 liters Buffer D (50 mM Tris- HCl pH 8.0 at 4°C), containing 2 mM 2—mercaptoethanol overnight. The above C50 eluate was applied to a DB 52 column (1.5 X 6.5 cm) pre-equilibrated with Buffer D. Elution was carried out with a linear gradient of Buffer D (100 ml) and 0.1 M NaCl in Buffer D (100 m1). Fractions of 3.4 ml were collected and analyzed for NTPH activity (See Section III. C. l. a.). Tubes containing maximal NTPH activity were pooled and concentrated by ultrafiltration cell with a PM 10 membrane. The concentrated DE 52 eluate was dialyzed overnight against 2 mM Tris-HCl (pH 7.4 at 4°C) containing 2 mM Zemercaptoethanol. 54 Granulated gel isoelectric focusing was performed with modifications by the method of Radola (73, 74). A gel slurry containing 2.5 ml Biolite 4-6, 7 pl 50 mM 2-mercaptoethanol, 47.5 ml H20 and 3.5 g Sephadex G-75-40 was deaerated, poured into a gel tray 1 X 11.7 X 20.4 cm, and allowed to dry to proper consistency. The dialyzed DE52 eluate was applied as one band on the plate. Electrofocusing was carried out using a Bio-Rad Model 1405 unit attached to a Lauda/Brinkman K-2/RS circulator to maintain the temperature at 4°C. The anode and cathode solutions were 0.2 N H2504 and 0.2 N NaOH, respectively, and Whatman #1 filter paper was used for electrode wicks. The electrofocusing was conducted at 150 V for two hours and 500 V for four hours. Sample strips of gel were collected via spatula and the protein eluted from the gel using 5 m1 disposable syringes as columns. A 50 mM Tris-HCl buffer (pH 7.4 at 4°C) containing 2 mM 2emercaptoethanol was utilized as the elution buffer. Samples were concentrated by ultrafiltration as before and assayed for NTPH activity. The preparation at this point generally contained two major protein bands on pdlyacrylamide gels (See Section III. C. l. e.), one of which was NTPH (as identified by activity staining), comprising approximately 40-50% of the total protein. In some preparations where NTPH was less pure at this point, the protein solution was 55 dialyzed against Buffer D and applied to a second DE52 column (1.0 X 4 cm). Elution of NTPH was carried as before, but each chamber of the gradient mixer contained only 50 ml solution. Additionally, in response to the results of Section IV. A.4, dithiothreitol ultimately replaced 2- mercaptoethanol in the purification procedure as a more effective sulfhydryl compound for the maintenance of NTPH activity. 56 IV. Results A. Physical Studies of Human Red Cell NTPH 1. Determination of isoelectric point by isoelectric focusing in sucrose density gradients a. Determination of pI of NTPH using human red cell hemolysates Isoelectric focusing in sucrose density gradients was performed as described in Section III. C. 1. c. In the light gradient solution, 1 ml of a 1:1 hemolysate (packed red cells lysed in an equal volume of 2mM DTT) replaced 1 ml H20. Following the focusing period, fractions were collected and analyzed for NTPH activity, absorbance at 280 nm, and measurement of the pH at 4°C. The results are shown in Figure l. The pI of human red cell NTPH as determined in this experiment is 4.65. b. Determination of pI of NTPH using partially purified human red cell NTPH Isoelectric focusing was performed and fractions collected and analyzed as above. The sample, however, was 100 p1 partially purified human red cell NTPH (specific activity 13,000 units/mg protein,l300 units/ml) prepared by the method of Morris (8) through the CM-Sephadex C-SO- 120 step resulting in approximately SOO-fold purification. The results of this experiment are presented in Figure 2. The pI of human red cell NTPH is 4.74 in this determination. 57 G) <3 0 O O O (D (D Q“ (“0!49014/94W") Rumor Hle (+) o o 9. O O “3 I .Oov um omsflfiumuoo mm3 mm was .usmeomuo muemsoo omouosm m as ten: mm3 m.m I v oDMHEhmem mm ouomhaoEom Hamo com o mo mmez mo mnemooom ownuomaoomH .H muomam Eooo _EN.V tense: cozoot . e 1 . 11-_)1l e -0. an O O o JQN O¢ O 00 . e ) o .. Emmet“... O x Z d en .00 .mx MW U” o 1% 1Q6 0 -2 new - — p — F — .Oev um oosHEstoo was no one .usowomsm wuemsmo omososm m as toms mos m.m I v ouaamEsmnm mm mmez Haoo com omemessm madneusmm mo mnemooom oesuooaoomH .N mesmem Eco» .Emo 835:: 5:03.". mm 0% mm) ON 9 o. O 0 IO 0 o 9 58 i/suun) Kumov HdiN(-°—) o o 9 31014901 0 O N . -vd - O 00 0 1Q0 .. so \I .. ..___\ . -N._ V .e W e 0 e o m.— 0 .uo 09 o QN 59 The results of these pI determinations from hemolysates and purified preparations of NTPH are in close agreement. The pI of rabbit red cell NTPH has been determined to be 4.3 - 4.5 (75) -— also quite close to that observed here for the human enzyme. These data present no evidence for the existence of isozymes of NTPH. 2. Further purification of human red cell NTPH Human red cell NTPH has beenppurified nearly 1200- fold to a specific activity of 25,700 units/mg protein by Morris (8). This preparation still contains several con- taminating proteins, however, further purification of NTPH was attempted in an effort to provide enzyme sufficiently pure for analysis of subunit molecular weight and other biochemical studies. The method of Morris was followed through the lysate, calcium phosphate gel, and CM-Sephadex C-50-120 steps which yields an NTPH preparation with a specific activity of about 12,000 units/mg protein: Additional purification steps, DE 52 ion exchange chromato— graphy, granulated gel preparative isoelectric focusing, and in some preparations, a second DE 52 step was carried out as described in Section III. C. 2. b. The first DE 52 Steps results in about 40% recovery of original NTPH activity with a specific activity of 26,000 - 76,000 units/mg protein in different preparations. The subsequent iso- electric focusing and DE 52 chromatography steps add little 60 to the specific activity of the NTPH preparation due to the considerable rate of loss of NTPH activity which occurs in highly purified preparations. However, analytical poly- acrylamide gel electrophoresis of aliquots of the purification steps auplicate gels are stained for protein and NTPH activity) reveals that isolation of NTPH is being achieved. The best NTPH preparations achieved by this purification scheme contain two major protein bands as shown on poly- acrylamide gels. Figure 3 shows polyacrylamide gels 6f an NTPH preparation after granulated gel preparative isoelectrid focusing and after the second DE52 step. The latter gel reveals two major protein bands of nearly equal staining intensity which comprise approximately 90% of the total protein present as measured by densimetric tracing in a Gilford 2400 spectrophotometer. The more anodic band of the two major bands corresponds to the area of NTPH activity staining of an identical gel. Studies of NTPH subunit molecular weight using SDS—polyacrylamide gel electrophoresis are presented in Section IV. A. 3. c. 3. Moledular weight estimations of human red cell NTPH a. Gel filtration of partially purified NTPH Partially purified NTPH prepared by the method of Morris (8) through the CM-Sephadexxc 50-120 step and purified human hemoglobin prepared by the method of 61 (+) Gal 0 Gel b Figure 3. Polyacrylamide Gel Electrophoresis of NTPH Purification Fractions The electrophoresis procedure as described in Section III. C. l. e. The gels have been stained for protein as described in Section III. C. 1. h. Gel a represents the NTPH preparation after the isoelectric focusing step. Gel b represents the NTPH preparation after the second DE52 step. 62 Winterhalter and Huehns (76) were applied to a Sephadex G—100 (medium) column and eluted with 50 mM Tris-HCl with mM GSH, pH 7.5 at 4°C. The NTPH activity eluted slightly ahead of the hemoglobin peak (Figure 4) which suggests a molecular weight in excess of 65,000 daltons. Gel filtrations of human red cell lysates with G-100 gave similar results. b. Molecular weight estimation of human red cell NTPH by sucrose density gradient centrifugation Sucrose density gradient centrifugation was performed as described in Section III. C. l. b. The sample was partially purified NTPH prepared through the CM-Sephadex C 50-120 step (specific activity 13,000 units/mg, 1300 units/ml) as described by Morris (8). Human hemoglobin and bovine pancreatic deoxyribonuclease were used as markers. The calculations were based on S values for the markers as reported by Chiancone gt gt. (66) and Lindberg (67), respectively. Figure 5 shows the results of analysis of NTPH, DNase, and hemoglobin-containing fractions of the gradient. Calculations based on the S value and molecular weight of DNase yield an NTPH S value of 2.97 and a mole— cular weight of 34,200. Calculations based on a graphic method (shown in Figure 6) which uses both DNase and hemoglobin parameters results in an33value of 3.00 and a molecular weight of 35,100 for NTPH. These measurements 63 91'? V (4») 8 I .DxOD 2H poneuomoo mm #50 oofluumo mmS sowuoupaflm How mesz upstaged asseeueee no eoeuepesea sew oesuo xmeeedem 28m _E 3855:: c3895. .wm- NN m. . .3 Io. m .v oppose OOON 000v 000m 000m (uouomysuun) KIIAIIov Hle 6+) 64 spun ascNG (4L) ID S2 9 coeummnmesusoo someomuw anemone mmonosm mmfiz MO 5095.5:— Sarah—ugl— VN DO 9% (+9 2 . ON/. I * .m mnsmflm 738:» e, s (IUJ/dll sanerMIIAIIov Hd1N(+) IQ fl Inc, '59 65 8 value ' )5 Distance from center of sample (in fractions) ..: 32 3. . B . E 3.! Ch 2 .30 Q , N 2.9 2.0 2.5 3.0 3.5 4.0 8 value Figure 6. Determination of S Value and Molecular Weight of NTPH A) plot of distance traveled in gradient versus S value B) plot of S value versus 2/3 log molecular weight 66 both differ substantially from the molecular weight estimate of 65,000 obtained by gel filtration. c. Subunit molecular weight estimation of human red cell NTPH by sodium dodecyl sulfate— polyacrylamide gel electrophoresis SDS-polyacrylamide gel electrophoresis of purified NTPH was carried out as described in Section III. C. l. f. Figure 7 shows the results of SDS gels of protein standards (gel a), purified NTPH ((gelbb) sample is same as that of gel b in Figure 3), protein band 1 ofgel b, Figure 3 (gel c) and protein band 2 of gel b, Figure 3 (gel d). The samples loaded on gels c and d were cut from Bis-acrylamide cross- linked gels, extracted from the gels and lyophilized as described in Section III. C. l. g. The sample of gel d is the protein band which corresponds to NTPH activity staining on non-denaturing gels. As can be readily observed, gels b, c, and d exhibit the same two stained bands. The mole- cular weights of these bands are 26,000 daltons and 40,000 daltons as estimatedgraphically from a plot of the log subunit molecular weight of the protein standards versus their relative mobility (Figure 8). It must be noted that more recent purification procedures utilizing dithiothreitol in place of 2—mercapto- ethanol have resulted in elimination of the 26,000 dalton ;polypeptide. This polypeptide is perhaps a proteolytic <=leavage fragment of the NTPH 40,000 dalton polypeptide. 67 BSA 68,000 PK 57,000 LDH 35.000 40.000 DNase 3|,ooo CHYA 25.000 26.000 Figure 7. SDS-Polyacrylamide Gel Electrophoresis of Purified Human NTPH SDS-PAGE was performed as described in Section III. C. l. f. and gels stained as described in Section III. C. 1. h. Gel a) protein standards as described in Section ILL. C. l. f. Gel b) purified NTPH after second DE52 step Gel d) protein band 1 of gel b, Figure 3 Gel d) protein band 2 oHTPH) of gel b, Figure 3 The migration of the tracking dye in each gel has been marked byidnsertion of pieces of copper wire. 68 (D'xl 00 0| 0 1 a 0 04 0 Molecular Weighl(xl0‘3) DO 0 g L L L l .2 .4 .6 .8 Relative Mobility Figure 8. Determination of Subunit Molecular Weight of NTPH The arrows mark the relative mobilities of the two protein bands of gels c and d of Figure 7. The corresponding molecular weights are 26,000 and 40,000. 69 It is proposed that NTPH is a dimer composed of identical 40,000 dalton subunits. The lower estimated molecular weight of 34,000 - 35,000 by sucrose density gradient centrifugation probably reflects a steric phenomenon. That NTPH is a dimer is suggested by the larger molecular weight presented by gel filtration. 4. Studies of the thiol requirements of human red cell NTPH. A recent report by Vanderheiden stated that human red cell NTPH does not exhibit a requirement for thiol compounds (7). This was based in large part upon the inability to demonstrate inhibition of NTPH activity by N—methylmaleimide or iodoacetamide. Since previous observations in this laboratory indicated that rabbit and human red cell NTPH require the presence of thiol compounds for maximal activity (1, 8) this question has been re-examined. The effect of thiol reagents upon NTPH was studied using N-ehtylmaleimide, iodoacetamide, and p-hydro- xymercuribenzoate (See Section III. C. l. j.). These studies utilized both human red cell hemolysates and a highly purified preparation of human red cell NTPH (specific activity-~475,000 units/mg; see Sections III. C. 2. b. and IV. A. 2.). Preincubation of each NTPH preparation with a range of concentrations of the three thiol reagents followed by 70 analysis for NTPH activity revealed marked sensitivity of NTPH to all three thiol reagents (Table 1). It can be noted that 10 mM concentrations of each of the three thiol reagents in the preincubation mixture produce unmistakeable inhibition of NTPH. In general, the more highly purified preparation of NTPH was more severely inhibited at a given concentration of thiol reagent (for example, note 5 mM N-ethylmaleimide, 20 mM iodoacetamide and 2 mM p-hydro- xymercuribenzoate). Analysis of possible effects of the thiol reagents upon the yeast inorganic pyrophosphatase utilized for inorganic phosphate production in the coupled assay system was conducted in assays from which NTPH was omitted and the ITP substrate replaced with 5 mM sodium pyrophosphate. Neither N-ethylmaleimide, iodoacetamide nor p—hydroxymer- curibenzoate addition at the highest levels present in the assay nor omission of DTT from the assay had any measurable effect upon the rate of inorganic phosphate production from sodium pyr0phosphate. Nor was any effect of the thiol reagents upon color development in the phosphate assay observed when thiol reagents were added to reaction mixtures containing the standard phosphate solutions. Hence no effects of the thiol reagents upon the coupled assay system used in the analysis of NTPH activity were occurring. TABLE 1. 71 Inhibition of NTPH Activity by Sulfhydryl-Specific Reagents Hemolysate NTPH Highly Purified NTPH % Inhibition by: Reagent Concentration* NEM IAA pHMB NEM IAA pHMB mM 1 8 0 0 23 0 30 2 - 0 8.1 41 0 91 5 36 0 97 43 0 89 10 51 12 100 59 ll 93 20 72 28 - 78 33 - 50 98 71 - 97 71 - *NTPH preparations containing 0.8 mM BME and TriSoHC1 (pH 7.4) were preincubated at 4° for 30 minutes in the presence of the reagents as indicated. NTPH was then assayed for enzyme activity in a 1.0 ml reaction mixture (with 2 mM DTT). A 50 pl aliquot of 72 The thiol requirement of human red cell NTPH has also been studied by examining the ability of BME, GSH and DTT to stabilize NTPH activity during prolonged dialysis at 4° C (Table 2). Following dialysis the NTPH preparations were analyzed both in the normal assay system, which contains 2 mM DTT, as well as in assays from which thiol compounds were omitted. Activity values were compared both with samples dialyzed in 25 mM Tris-Cl, mM DTT during the dialysis period (No. 5 of Table 2). These studies revealed some rather unexpected effects of certain thiol compounds upon NTPH. In particular, it became apparent that BME in the dialysis medium led to an increased loss of NTPH activity. Activity values observediin the sample dialyzed against DTT, and assayed without added DTT, were consistently lower than those observed when no thiol compound was added to the dialysis medium. This phenomenon was apparent with both the partially purified C-SO eluate NTPH (compare 1 and 2 of Table 2) as well as with crude hemolysate NTPH (compare 6 and 7, 8 and 9, Table 2). The partially purified NTPH samples were particularly sensitive to oxidation during dialysis in the absence of thiol as well as in the presence of 6MB and GSH. On the other hand, inclusion of DTT in the dialysis medium led to activity values similar to that seen with the non-dialyzed sample kept in DTT (compare 4 and 5 of Table 2). TABLE 2 . Upon 73 Effect of Selected Sulfhydryl Compounds NTPH Stability Source and Treatment of NTPH Assay Conditions No DTT Added With 2 mM DTT units/m1 units/ml l. Dialyzed* C-50 column eluate*** NTPH 160 830 2. Same as (1) plus mM 8MB 70 950 3. Samse as (1) plus mM GSH 170 980 4. Same as (1) plus mM DTT 670 2,000 5. No sialysis plus mM DTT 660 1,800 6. Dialyzed* hemolysate Sp; act. sp. act. (SAF)*** 25 36 7. Same as (6) plus mM BME 0.6 37 8. Dialyzed hemolysate sp. act. sp. act. (LJM)** 35 51 9. Same as (8) plus mM BME 11 50 *The C-SO column eluate was dialyzed at 4° twice against 500 volumes 25 mM Tris-HCl (pH 7.4) for a total of 15 hours with the additions as indicated. **See reference 8. ***Freshly prepared hemolysates analyzed in the presence of 2 mM DTT possessed specific activity values of 38 for SAF and 50 for LJM. 74 When analyzed in the presence of 2 mM DTT, the dialyzed samples showed significantly higher NTPH activity than in the absence of added DTT. Hemolysate NTPH and the partially purified NTPH dialyzed against mM DTT possessed full activity in the presence of 2 mM DTT. Hemolysates dialyzed without thiol or in the presence of BME showed no inactivation when analyzed with 2 mM DTT in the assay medium. In order to distinguish between possible inactivation of NTPH during dialysis, inactivation of NTPH during the time course of the incubation assay and possible re- activation by DTT in the assays containing DTT, the kinetics of NTPH activity from the partially purified NTPH preparation were studied following dialysis (Figure 9). Analysis of NTPH activity in an assay lacking DTT revealed complete inactivation of the enzyme during the course of the 20 minute incubation period, regardless of the composition of the dialysis medium. On the other hand, assays containing DTT were linear for the entire period. It is_noteworthy, however, that the sample dialyzed against mM BME and analyzed in the presence of DTT possessed only 60% of the NTPH activity present in the sample dialyzed in the presence of mM DTT. Dialysis in BME resulted in loss of NTPH activity which was distinct from the inactivation which occurs during incubation without DTT. 75 .Amosomem sumo one mmsHH soxounv Hosea popes msexoma madame cw Ho AmmHsmHm oflaom tee emcee eesomv see as N meeeeeueoo 1.e .H .o .HHH soauoom moms whammm HE o.H osmosaum one as >ue>fluom damned maesHaEwH Mom ouammmm ouo3 muosoeam a: om one .Amoamsmwnuv mzm SE mo oosomoum any ca so Amoaonwov BBQ 25 m.o mo mosmmoum mop aw Hugues .usmflsso>o emusseee mes impeded omuov meez openness seseeusee monsooaoo Hmsomsmasm mo mundane Ho accommnm as coeumnsosH msHHDQ huw>fluod mmBz .m shaman 76 no monomosm as mosoomsoo Hmuohnmaom mo mosomne cosponsosH msflsso zuw>wuom mmez .m anomwm 32:55 as: 0.0 J l o (salouI If ) aaznoaa/IH dll 77 Partial reactivation of inactive NTPH with DTT could be demonstrated by addition of DTT to a preparation incubated in the absence of DTT until complete inactivation had occurred followed by the addition of DTT and continued incubation (Figure 10). Following an initial reactivation period the slope of NTPH activity between 30 and 40 minutes of incubation indicated 78% recovery of the NTPH activity present in the control incubated with DTT throughout the entire 40 minute period. While the biochemical details of the inactivation- reactivation process which occurs with NTPH require much more detailed analysis, data presented here support the conclusion that human red cell NTPH has an absolute require- ment for the presence of a reduced thiol compound, particularly in purified preparations, and that DTT is the thiol compound of choice. Inactivation brought about by the absence of a thiol compound in the reaction mixture is not completely reversible. No explanation can be offered for the reported inability to inhibit human red cell NTPH with thiol reagents or omission of thiol compounds (7). 78 .oumnnmonm osnmmnone now oonmamna one ooumoeone moownom mafia on» Do mnoeumnnone m>fluommmon onu Eonm om>OEoH mnm3 “HE o.av muonowan .mounnea om HonOHDHoom no How omnnwunoo mm3 nomo mo noeumnnonfl one Amosmnom oHHOm .onwa owaomv noflunomnmno 90 SE N on oooom was BBQ one oooe>wo ma3 nOflumnnonH nausea on» mounnwa on an .Amonea noxomn one mmaoneo nooov mounnee om How BBQ mnwxoma EnHowE hmmmm nfl ooumnnonw mmz nowpnom onooom n .Ammaoneo oeaom one onHH oHHOmV noeumnnone one mo mnflnnemon onu um oooom BBQ 28 m manHounoo EnHooE wanna nH owumnnonfi mo3 nOeuHom mno .mze es .Aoe um e.e met so mess as mm amended page: upe>o eeusseee mes impeded emuov meez edemehee enamepeee HoufloHnDOHnuwa mo mmBz ooum>fluomnH mo noHum>Huomom .oa apnoea 79 Houwmnnuoflnufla an mmez emue>nuoecH mo cofiue>fluommm A $555 V ms: 06 Om ON A: .oa munmwm o QIldIlIQIIIPIOIIIOIIII d 0 PO‘ 0 O “? C) O. “? Q N “2 a: Q N) (salow d) GBZA'IOHGAH dll 80 B. Population Surveys of NTPH Activity in Human Red Blood Cells 1. NTPH activity in a normal population In early studies of human red cell NTPH in this laboratory, a marked variation in the level of NTPH activity among individuals was observed. To examine the extent of this variation and to provide a framework for further bio- chemical and genetic studies of NTPH variability, a population survey of NTPH activity in human red cells was carried out. It must be noted that prior to the study presented here, a population survey of "ITPase" activity in human red cells had appeared (4). The "ITPase" activity measured was likely the result of the coupled activities of NTPH and endogenous inorganic pyrophosphatase. Thus, the measure- ment of the Pi produced by the coupled reaction may be underestimated if sufficient inorganic pyrophosphatase is not present to cleave the PPi produced by NTPH. Other conditions of the assay system used were suboptimal, particularly the pH at which the incubation was conducted. Furthermore, the subjects for the population survey were drawn from a mixed population without regard for racial origins. Population data on human red cell galactokinase which show vastly different mean activities for black and 81 white populations point out how important considerations of racial origin can be in population surveys (57). Serious reservations, therefore, are entertained about the "ITPase" population data. In our study, random blood samples from a Caucasian population were obtained from the American Red Cross and the erythrocytes analyzed for NTPH activity over a two month period. No information on sex of the donors was received. The results for 262 individuals are shown in Figure 11. Two distinct modes of activity are readily apparent. A low NTPH group, comprising 17.2% of the sample population, ranges in specific activity from undetectable levels to 25 nmoles ITP cleaved/20 minutes/mg hemoglobin. The high NTPH group comprising 82.8% of the sample population ranges in specific activity from 25 to 125 nmoles ITP cleave/20 minutes/mg hemoglobin. Thus, variation in NTPH activity in red blood cells from the high "normal" range is quite common. These results are comparable to those of the "ITPase" population survey presented by Vanderheiden (4). His data also resolved into two classes of activity, a low group, 19% of the population, and a high group, 81% of the population. However, more recent population data presented by vanderheiden and Zarate—Moyano (l9) reveal a unimodal distribution of NTPH activities. The assay method utilized in the later study has the same faults described above. The lack of concordance in his data is unexplained. 82 .mnOHuHonoo wemme oueoneum noon: oohm ue mounnwa om nH mBH oeosn H mmuwaouown nownz meannm onu mo unDOEe uenu me omnwmoo we aufl>fluoe mmez mo pens e .mueuumnsm may we maH nun3 mnOHuHonoo hemme oueoneum Moon: mueowamwuu nfl oomhaene me3 Hmsefl>flecn some mo mummnfl Hamo emu e no sun>nuom mmez may noaueanmom neHmeoneU e mo mHHoU own on» nH mufl>nuo< unmeomdm mmez mo sm>usm Eocene e .HH musmnn 83 noflueanmom neflmeoneo e no maaou omm onu an mafi>fluo< onenomdm mmez mo >m>usm eoecem e .HH musmnn A=_no_mosmz ma\me_cav sn_>_nu< u_e_uoam Ian: ON. 0.. oo_- on on om om oe on om o. - o LEL _ .I 1 I 4 A i .l _ 1 T I I... r . III 4 N— mp om em mm mm (slenp;A;pu1) KauanbaJJ 84 To further analyze the population distribution of red cell NTPH activity as presented in Figure 11, normal distributions were fit to the two groups of NTPH activity by a method described by Croxton (77). The normal distri- bution for the high NTPH data was calculated excluding the four individuals with specific activity above 95. The resulting normal distributions are presented with the population distribution of NTPH in Figure 12. The goodness of fit of the data to the calculated normal distribution was examined using the Chi-Square test. The high NTPH population showed adequate fit to the normal distribution (xfil = 7.754, .5o>nnm noHueHnmom mmBz ou uflm mnowunnfluumfla HeEuoz 38.822 sweetest/tang 0.28% :92 ON. -o__ 8.8 om on om on- ow on 8 0. .NH musmnn , O [EL '4 r , x on e- C) (D Cd .. N (slonp wpug ) Kouanbald 981% 0d N5 86 2. NTPH activity in the red cells of patients with paranoid schizophrenia and schizophrenia Blood samples of eight paranoid schizophrenic and four schiZOphrenic psychiatric patients were obtained from Dr. R. Kobes, Department of Clinical Psychopharmacology, National Institutes of Health, St. Elizabeth Hospital, Washington, D.C., presently of the Department of Biochemistry, Michigan State University. The small number of samples available did not differ substantially from the activity found in the general population (Figure 11). Two samples of twelve (16.7%) total patients were in the low NTPH range (specific activity 25 units/mg hemoglobin), while among paranoid schizophrenics alone two of eight (25%) were in the low NTPH range. The NTPH specific activities of the psychiatric patients are listed in Table 3. C. Studies of Biochemical Parameters of NTPH Variation 1. Isoelectric focusing in polyacrylamide gels The variation in NTPH activity in red blood cells of individuals is not associated with electpophoretic variants (2, 15). However, the technique of isoelectric focusing in polyacrylamide gels has proven to be more sensitive than electrophoresis for detection of heterogeneity in protein samples. Hemolysates from individuals with NTPH specific activities ranging from 3 to 65 were applied to 7.5% polyacrylamide gel rods containing Pharmalyte 4 - 6.5 TABLE 3 . Patient 1 10 ll 12 87 NTPH Activity in the Red Cells of Psychiatric Patients NTPH specific activity Diagnosis units/mg hemoglobin Paranoid Schizophrenia 65 " 67 " 3 " 53 " 69 " 50, " 58, " l7, Schizophrenia 48 " 52, " 69 ll 67 39 58 15 40, 38 For patients 6, 7, 8 and 10, multiple samples were analyzed for NTPH activity. duplicate analyses are recorded above. The results of these 88 The electrofocusing was carried out and the gels were stained for NTPH activity after incubation at 37°C with ITP and the normal NTPH reaction mixture. One stained region appeared on gels incubated with ITP while gels incubated with ATP in place of ITP gave no staining reactions. The intensity of the stained region varied from gel to gel in proportion to the NTPH specific activity applied to each gel. No differences were observed other than those of staining intensity between gels containing the focused hemolysates of individuals with high and low NTPH specific activities. When hemolysates of high and low NTPH individuals were mixed and applied to one electro-focusing gel, only one staining band was observed and its intensity was inter- mediate to those of the high and low NTPH gels alone. Thus, the quantitative variation in NTPH specific activity among individuals is not associated with differences in the isoelectri3focusing properties of their respective NTPH molecules. 2. Substrate specificities of red cell hemolysates The ability of hemolysates of low and high NTPH to hydrolyze substrates other than ITP was studied in an effort to determine whether differences in utilization of other substrates could be observed in the two groups. Purified' rabbit red cell NTPH utilizes dITP at a similar rate to ITP and utilizes XTP only slightly less (1). Human red cell m 89 NTPH when partially purified reveals a pattern of substrate utilization similar to that of the rabbit red cell enzyme except that GTP is hydrolyzed at only 2-3% of ITP by human NTPH while rabbit NTPH hydrolyzes GTP at a rate 10% that of ITP. Table 4 presents the results of utilization of the substrates ITP, dITP and XTP (all at concentrations of 0.5 mM in the NTPH reaction mix) in hemolysates of 2 low (LLF and DD and 2 high (SAF and AJM) NTPH subjects. It is apparent that the specific activities of the hemolysates with these substrates tested follow the same general pattern. When the spbgéfic activities of the hemolysates using dITP or XTP as the substrate are expressed relative to the specific activities with ITP as substrate, the low (LLF and DY) and high (SAF and AJM) subjects express similar degrees of substrate utilization. The somewhat higher relative activity of subject DY's hemolysate with dITP as the substrate is not statistically significant using the t-test (t = 4.078, 0.5nuom mean non " souuom o.mm u >ua>fluoe oflmsoomw mmBz .mnm pooflnnm Am mH.H.I nHQonoEwn mou " Eouuom mvv.l wuw>fluoe mo» u Eouuom m.vH u mufl>fluoe UHMHoomm mmBz .>A> noanSm Ad .mnOHuoeum xwm mo noeo MOM omnHEHmuoo euoz mpemwa mua e mo nofiueuunoonoo nHQOHmOEon one >9H>wuoe mmaz Aomnv muflmneo >3 nowuenemom HHoU oem .mH onnmflm 93 oil 23». _E\=_no_ooEez an. M3333233fi43fl3w . q 4‘ q q u u A J / 5% ,wllw 1&1 T013885»; 325333 2:88 1&2 20 4O 60 80 Position in gradient(%from top) Red Cell Separation by Density (Age) Figure 13. 94 TABLE 5. Red Cell Separation by Density (Age) NTPH Specific Activity (Units/mg hemoglobin) Subj. Non-separated Top 10% Bottom 10% Bottomztop red cells of gradient of gradient activity ratio LLF 3.0 3.2 1.4 .438 VLV 14.8 18.5 8.2 .443 WGC 24.2 28.0 12.0 .429 HPH 26.5 33.5 12.4 .370 SAF 32.0 41.0 20.4 .498 CV 37.5 47.8 24.5 .513 AJM 52.8 73.2 32.4 .443 JDC 71.7 107.1 42.5 .397 DL 101.2 127.8 67.0 .524 95 the range of NTPH activities studied. The youngest red cells of the individuals with low NTPH activity display correspondingly low activity while the oldest red cells display activity in a ratio to the youngest cells which is similar to that seen in individuals with high NTPH activity. These data present evidence that no differences exist in in vizg stability of NTPH in the red blood cells of low and high NTPH individuals. 4. Thermostability of NTPH The thermostability characteristics of NTPH in hemo- lysates were studied in an effort to demonstrate a qualitative difference in NTPH variation in addition to the observed quantitative variation. Hemolysates from individuals representing a wide range of NTPH specific activities were incubated at 65°C for 12-15 minutes and analyzed for loss of activity with time. The incubation temperature of 65°C was chosen for its utility in separating phenotypic classes of thermostability and for the modest amount of time needed to reduce enzyme activity to 50% of the activity of the original-sample (approximately 7 minutes in high NTPH subjects). Figure 14 shows examples of thermostability data of two subjects and of a mixture of the hemolysates of the two subjects. The hemolysate of subject LLF, with an NTPH specific activity of 4.4, exhibits a greater degree of lability at 65°C than does the hemolysate of AJM, with an NTPH specific activity of 82. The sample exhibiting inter- Activity as Percentage of Unheated Sample IOO 90 80 7O 60 50 4O 96 T I l l 1 (D F- O 2 4 6 8 Time (minutes) Thermostability at 65°C of Hemolysates of two Subjects Figure 14. Circles, AJM (Sp. Act. 82), Squares, LLF (Sp. Act. 4.4) and Triangles, LLF:AJM, 15:1 (Sp. Act. 10). 97 mediate lability represents an approximately 1:1 mixture of NTPH activities from hemolysates of subjects LLF and AJM, requiring a 15:1 mixture of the hemolysates due to the large difference in specific activities of the respective lysates. The intermediate thermostability characteristics of this mixture of hemolysates indicates that the decreased thermostability of the low NTPH sample compared to the high NTPH sample is not due to intracellular effectors, but rather is apparently an inherent characteristic of the NTPH of each individual. Other mixing experiments of this sort corroborated the data presented in Figure 14. Figure 15 presents the data of Figure 14 as the log of NTPH activity versus time. The activity scale has been arbitrarily fixed and each set of data adjusted to 100% activity at the same point so that the slopes of the activity loss with heating can be readily compared. Lines have been fit to the data and slopes calculated using a computer program of the least squares method. The slopes of the loss of NTPH activity with heating of each sample,eexpressed as log nmoles ITP hydrolyzed/10 minutes (tQSAE.), were as follows: LLF, -.800 (i .045), AJM, -.438 (i .009), and the mixed sample, -.552 (i .008). Pairwise tests of difference between the regression coefficients (slopes) using the t-test yielded in all cases P<0.01 (78, 79). Log nmoles of ITP Hydrolyzed ~03 - 0.4 - 0.5 '06 -O.7 98 It 0 In .1 H I O I. . -‘ A O O A I (- — O p- q E] l I g A l l l O 2 4 6 8 IO l2 TIME (minutes) Figure 15. Semilog Plot of Data of Figure 14 The lines have been fitted to the data by a computer program using the least squares method. Data points represent the means of triplicate assays. The lines have been adjusted to the same initial point such that the ordinate is scaled starting arbitrarily at 0 for each sample. 99 The differences seen in thermostability of NTPH from high and low activity hemolysates as described above are substantiated by the data presented in Table 6. Figure 16 depicts the data of Table 6 graphically. Multiple experiments were performed with hemolysates of some subjects over a period of several months with consistent results. Three thermostability phenotypes can be discerned, one with NTPH specific activity below 5, a second with NTPH specific activity between 12 and 25, and a third with NTPH specific activity above 25. These groupings correspond well to the population survey data presented in Figure 11 and offer further evidence (in addition to the poor fit of a normal distribution to the low NTPH mode of thep population distribution, Figure 12) that the low NTPH region (NTPH specific activity of <25) is actually composed of at least two groups. It must be noted that the absence of data from subjects with NTPH specific activities between 5 and 12 leaves some ambiguity concerning the boundaries and overlap of these two thermostability phenotypes in the low NTPH region. The thermostability slopes (expressed as log nmoles ITP hydrolyzed per 10 minutes) of the 0 - 5 NTPH specific activity hemolysates ranged from -0.738 to -1.405, the slopes of the 12 — 25 NTPH specific activity hemolysates ranged from -0.469 to -0.606, and the slopes of the <25 NTPH specific activity aanged from —0.391 to -0.471. It will be noted that Table 6 lists duplicate mmuanE 00\oonmaouomn mBH 00008: 00H we oommoumxo we emoam wuflaflneumoEHonu one 100 _ A000.0 «0000.0: 0AA AH00.0 «0000.0: A000.0 “VH00.0: A000.0 «0000.H: Mo A.m.m Hy nemz A.m.m H0 new: A.m.m «V nemz 000.0: 0.00 mo 000.0: 0.00 >0 000.0: .0.00 mm 000.0: 0.H0 >0 000.0: 0.0m 0m H00.0: 0.00 000 000.0: 0.00 03 000.0: 0.00 qu 000.0: 0.00 000 000.0: 0.00 ooh 000.0: 0.H0 20¢ 000.0: 0.00 hem 0H0.0: 0.0a 00 000.0: 0.00 204 000.0: 0.00 000 000.0: 0.0a 00 000.0: 0.00 20¢ 000.0: 0.00 000 000.0: 0.0a ow 0H0.0: 0.H0 20 000.0: 0.00 mmm 000.0: 0.00 :00 000.0: 0.0 0AA 000.0: 0.00 on 000.0: 0.00 max 000.0: 0.0H >q> 000.0: 0.0 mag 000.0: 0.00 000 000.0: 0.00 0mm 000.0: 0.00 >A> 000.H: 0.0 0AA 000.0: 0.:0 009 000.0: H.0m mm 000.0: 0.00 0H0 000.0: 0.0 an 000.0: 0.00 020 000.0: 0.00 mm 000.0: 0.0a mm 000.0: 0.0 No 000.0: 0.00 mm H00.0: 0.00 003 000.0: 0.00 mm 000.H: 0.0 we 000.0: 0.00 on 000.0: 0.00 003 000.0: 0.00 am 000.0: 0.0 an 1 omon uoe mm mane omwam #oe me wane omon uoexmw wane immoam uoe mm 0950 A0~nn0 00. Amaunv 00:0 Abunv 0:0 AnHQOHmOEmn mE\muwnnv muw>wuo¢ 00000090 mmBz 0.00 pm mundaneumosumne 0cm sufl>fluo0 00006000 0092 .0 00049 101 Thermostability slope (log NTPH units/lOminutes) 0 0 -l.4 '- I .0 .<'> .C'>.c'> .<'> C Figure 16. lb “2‘0 3‘0 4'0 50 6'0 76 8’0 NTPH SPECIFIC ACTIVITY(units/mg Hb) NTPH Specific Activity and Thermostability at 65°C S‘i: 1C1QtlfiiuflU§DIxibflCQW 102 thermostability determinations for several individuals. For the calculation of means and standard errors of each group, these duplicate samples have been averaged and used as one value. Since only two individuals in the 0 - 5 range were available for study, Table 6 lists the means and standard errors of the thermostability experiments performed with multiple samples of each of these persons. The mean slopes (i 8.3.) of the DY and LF samples and the 12 - 25 and greater than 25 NTPH specific activity groups were -1.052 c: 0.258), —0.857 (t 0.156), -o.551 (i 0.038), and -0.433 (t 0.021), respectively. The differences between the means of the three dis- cernable thermostability phenotypes were all highly significant using the t-test (78, 79), with all three pair- wise combinations. The results of the statistical tests are as follows: for the t-test of difference between the means of the DY sample and 12 - 25 groups, t = 6.16 with all degrees of freedom (P<.001), for the t-test of difference between the means of the LF sample and 12 - 25 group, t = 5.89 with 10 degrees of freedom (P<.001). Obviously samples DYand LF were significantly different from the above 25 NTPH group in thermostability also. Table 7 expresses the thermostability data in terms of time of incubation at 65°C required for 50% inactivation of NTPH. The data in this form once again clearly show the differences in lability of NTPH in hemolysates of individuals. 103 TABLE 7. Half-life Of NTPH at 65°C NTPH Time (minutes) for Specific Activity # Expts. 50% Inactivation (i S.E.) 0 - 5 7 3.26 (i .69) 12 - 25 12 5.55 (i .41) > 25 28 6.98 (i .39) 104 The evidence presented here of altered thermostability of NTPH in individuals with low NTPH specific activity marks the first physical parameter found which varies among persons with different NTPH activities. The usefulness of these data in defining the mode of inheritance of NTPH variability will be discussed later in relation to the family studies. D. Family Studies of NTPH Variation Preliminary studies of NTPH specific activity in family members had indicated that activity levels are not randomly distributed in these families, as observed in the population distribution, but instead were generally similar among family members. This observation suggested genetic control of NTPH activity in individuals. In his study of "ITPase" activity (4), Vanderheiden presented pedigrees of seven families, assigned genotypes on the basis of “ITPase" activity and suggested a one gene-two allele system of inheritance involving codominance of the alleles. Certain aspects of the study, however, cause reservations about these data. First, as previously mentioned in connection with the "ITPase" population distribution presented in the same paper, the assay conditions used were suboptimal for the measurement of NTPH activity. Secondly, the seven pedigrees presented contain much unusable data, as in several cases one or both parents of a particular family unit were not tested for "ITPase" activity. Also, the 105 assignment of genotypes appears to have been done somewhat arbitrarily, as overlapping of activities with respect to genotype are observed. In two cases, genotypes assigned to individuals are incompatible with the genotypes of other family members. From our observations it was determined that a family study of NTPH specific activity in red blood cells could be profitable in more accurately defining the mode of inheritance of the variation of NTPH specific activity. The data of the population survey of NTPH activity provided a basis for the formulation of test hypotheses concerning the inheritance of NTPH activity. Since the population data of Figure 11 exhibit two well-defined activity ranges, one can calculate gene and genotype frequencies for a simple one gene-two allele system of inheritance. The two alternative hypotheses for this simple system are: 1) that low NTPH specific activity is recessive, the individuals in that region being homozygous for a low activity allele, while the high NTPH specific activity region consists of heterozygotes and homozygotes for a high activity allele, or 2) that high NTPH activity is recessive, being homozygous, while the low activity region is composed of heterozygotes and homozygotes. For Hypothesis I, the gene frequencies of the low and high alleles, based on phenotypic frequencies of .072 for low NTPH and .828 for NTPH, are .415 and .585, respectively. 106 For Hypothesis II the gene frequencies of the low and high alleles, calculated in the same manner, are .090 and .910, respectively. To test the validity of these hypotheses, the pop— ulation ratio method was used as described by Li (80). This method involves the use of "complete selection" (60) where families are chosen at random for their NTPH phenotype. From calculated gene frequencies for each hypothesis, the expected frequencies of children of low and high NTPH activities can be obtained for each of the three types of parental matings; high NTPH x high NTPH, high NTPH X low NTPH, and low NTPH X low NTPH. The predicted phenotypic frequencies for the children of the various parental pairs for each of the simple one gene-two allele hypotheses are summarized in Table 8. NTPH specific activity was analyzed in the red blood cells of 42 Caucasian families selected at random for their NTPH phenotypes. These data are shown in Figure 17. An additional 15 families obtained from J. F. Henderson's laboratory at the University of Alberta, Edmonton, Alberta, Canada, were ascertained through propositi whose red cells synthesized high or low levels of I14CJ ITP from 114C] hypoxanthine, such that they are not a random sample for NTPH activity. The propositi were excluded, then, from zhese families when the previously mentionedgenetic 107 TABLE 8. Test Hypotheses for Inheritance of NTPH Variability (based on data of Figure 11) Phenotypic frequencies low NTPH .172 Parental mating type high NTPH .828 high NTPH x high NTPH high NTPH x low NTPH low NTPH X 10W NTPH Hypothesis I. Low NTPH as recessive Freq. Freq. Freq. Freq. Freq. Freq. Freq. low NTPH = q2 = .172 low Allele = q = .415 high Allele = l - q = p = .585 heterozygous high NTPH = 2pq = .486 homozygous high NTPH = p2 = .342 recessive children of dom X rec parental type recessive children of dom X dom parental type .686 .285 .030 .293 .086 Parental—type Freq. low NTPH children expected HIGH X HIGH 0.086 HIGH X LOW 0.293 LOW X LOW 1.000 NTPH specific activity (units/mg hemoglobin) Genotype(s) 0-25 > 25 fin NN, Nn TABLE 8. (Cont'd) Hypothesis II. High N 108 TPH as recessive Freq. Freq. Freq. Freq. Freq. Freq. Freq. high NTPN = q2 high Allele = q low Allele = p heterozygous low homozygous low N recessive childr recessive childr .828 .910 l - q = .090 NTPN = 2pq = .164 TPH = p2 = .008 en of dom X rec parental type = .476 an of dom X dom parental type .227 Parental type Freq. low NTPH children expected HIGH X HIGH 0.000 HIGH X LOW 0.524 LOW X LOW 0.773 NTPH specific activity (units/mg hemoglobin) 0-25 > 25 Genotype(s INN, Nn nn 109 .HIHHIQ ma Hounmneo umuflm Hon maflnz ~:H:o me on oonuomou me a hafleem mo HonuOE onu .msna .noHueHonom e nfinua3 ooeam onfluenmfimmo Henoenn aflneud ne one maenwann neEom n0 Honenn noflueuonom e .nOfluenmflmao Hopped waflEem onu moNHHHun nofins Eoummm mnemonenn moeam :oounu e ha axe» onu n0 on oouuowou on Haas mHenoH>HonH O D .maonE>m onu ooflmnfl ooouooou me one ooumaene me3 Henow>wonw noem mo muemwa Haoo ooh e we wufl>fluoe aflmwoomw mmBz oHeEom ooumou uon u an oomeoooo I Q u no oHeE moflaflfiem omuooaom haeoonem mo mufl>fluo< oamwoomm mmez AOIev 0H whomflm 110 Figure 17 a. 111 ji¥jfm 69 fig” Id 91- :a— r—EJN >I XI I®_ a—a Figure 17 b. 1: mceiving transfusions 112 Figure 17 c. 113 .00 ounmflm n0 oonwmoo one oomn maonE>m ooumwoom .nofiuennonfl mnflnno maaoo ocean ooh on» n0 onflnunexomxn HO0HH Eoum mBH no Hg 00 mHo>oH 300 no nmfln ooueasfinooe 0:3 Amsouue 03 oowmwunooflv Huwmomonm nmnounu oonfleunoome onoz mowHHEem omone moHHHewn emoomawm maeoeemmunoz mo moe>eeoe 60006000 0092 .0a mesmem 1‘ N m5 114 hypotheses were tested by the population ratio method. These 15 families are shown in Figure 18. The NTPH analyses of these 15 families were performed by V. Verhoef. Figure 19 shows the distribution of red cell NTPH activities of the parents of the 42 randomly ascertained families. The parents represent a random sample of NTPH activities. Of the 84 subjects, 19 were in the low NTPH range (22.6%). The low and high NTPH frequencies among this sample do not differ significantly from the population data of Figure 11 (x2 = 1.773, 1 df, o.s>p>o.1). Thus, analysis of the family study data using the population survey data as the basis of the population ratio calculations should be accurate. The distribution of NTPH activity of 223 family members of the 42 randomly selected families is shown in Figure 20. This distribution reflects that of the parents in that a similar proportion (49 of 223, 22.0%) of the individuals exhibit low activity. The pattern of this distribution suggests that a further mode of activity may be discernible in the high NTPH region between 25 and 50 units specific activity. This additional group will be discussed in more detail later. Figures 21 and 22 present the data of Figure 20 separated by sex. The male and female distributions of NTPH activity are quite similar. The proportions of low 115 0a ounmflm mo munonem mo 0u0>0uo< oameoomm mmez .00 onsmwm 052822 0535;0260 omomem 0&2 00. 0m 00 0.0 00 on ow om ON 0. O 0 v T I I 4 91 co 1 Q (slonpgAgpug) Kouanbaig 116 NH musmflm mo mHmQEmz >HHEmm mo wufi>flpod owmfiommm mmez .om wusmflm 38.852 ghee Ezpoq 07.8% 2&2 oo. om om ON ow Om 9» 0m ON 0. o ‘, O E (slonpgAgpul }O JeunnN) Kauenbag 117 ha wnswwm mo mwamz mo mufi>fluofl UHMHommm mmsz 232822. oe\£_§;:>:o< 2&0QO 1&2 .Hm‘musmflm m Ow. OmRON 0. oo . i. ow. Cum,om ON 00 O .% Ir “- .L 1, S! (0 4 $2 (smnpmgpuyMouanbaH 118 ha musmflm mo mmamsmm mo mpfi>fluo< oHMHommm mmaz .mm musmflm 38.852 3&3 CE»: o...=omn_w E5 8.. cm 8 E 8 on 8 on ON 0 o PC. L. I: H0 (slonpgAgpuy)KouanbaJ;1 119 NTPH individuals among males and females are 23.1% and 20.8%, respectively. The mean NTPH activities (: S.E.) of the distributions by sex were as follows: male, low NTPH -- 11.80 (i 7.07), high NTPH -- 60.93 (i 16.15), female, low NTPH -- 14.32 (t 6.69), high NTPH -- 62.15 (i 14.37). These values compare well to those of the population survey where the mean (iS.E.) of the low NTPH group was 14.28 (15.79) and that of the high NTPH group was 60.88 (i 18.30). These data offer no evidence that sex-related factors play a role in the determination of NTPH activity. The family data has been tabulated in Table 9. In pedigrees where more thancone sibship has been analyzed for NTPH activity, only one has been included in this tabulation to avoid bias in the sample. In each case of multiple sibships in one family, the youngest sibship has been used in Table 9 as a matter of convention. For each of the non-random pedigrees, the propositus has been excluded from the data entered in Table 9. The frequencies of the parental mating types of the randomly selected f families were high x high -- .595, high X low ~- .357, and low.x low -- .048. These frequencies do not differ statistically from those expected based on phenotypic frequeneiss of the population data of Figure 11 (X2 = 1.628, 2 df, .5>Pfi.4). 120 TABLE 9. Pedigree Data of Families of Figures 17 and 18 Offspring High Low Parental Pairs NTPH NTPH High NTPH Freq. of pheno- type of offspring Low NTPH I High NTPH x High NTPH Families A,B,C,E,?, FIGIHIIIKINIQIRI U,X,Z,AA,BB,DD, 77 2 EE,FF,GG,JJ,KK, LL,NN,Bo,Li,Sa, So 25 random + 4 non- random l O 975 .025 I High NTPH x Low NTPH Families D,J,L,M, O,P,S,V,W,Y,CC,HH, II,MM,PP,An,Bu,Zo, Be,Ca,Ke,Me,Mc,Wa, 36 29 Pu 15 random + 10 non-random .554 .446 1 LOW NTPH X LOW NTPH Families T,OO l 6 2 random .143 .857 Low NTPH activity is defined as specific activity below 25 units/mg hemoglobin High NTPH activity is defined as specific above 25 units/mg hemoglobin 121 Examination of the data in Table 9 reveals that neither simple one gene-two allele system of inheritance of NTPH variation as detailed in Table 8 is compatible with the observed results. Hypothesis I, with low NTPH as recessive, predicts that low NTPH X low NTPH parental pairs should have only low NTPH offspring, yet a high NTPH child (T-III—l) is recorded in one of the two families of this type. Conversely, Hypothesis II, with high NTPH as recessive, predicts that high NTPH X high NTPH parental pairs should have only high NTPH offpsring. One such family (Family H, Figure 17a), however, has two children with low NTPH levels. Because each of these cases relies on only one or two offspring to refute a hypothesis, a more accurate assessment of the utility of either of these simple 2 allele hypotheses can be obtained by examining the offspring of high NTPH X low NTPH parental pairs. Use of the chi-square test comparing the observed phenotypic frequencies of the offspring of high x low parents with the predicted frequencies calculated by the population ratio method in Table 8, reveals a highly significant difference between the observed results and the expectations of Hypothesis I. The differences between the observed family data and the expectations of Hypothesis II are not significant. The chi—square test for Hypothesis I with and withOut the Yates correction for sample size below 50 gave values of 7.43 and 6.71, respectively. 122 With 1 degree of freedom each x2 value indicates P < .01. For Hypothesis II, the chi-square values were 1.605 and 1.306, both resulting in P > 0.2. Hypothesis I, stating that low NTPH activity is a recessive trait, can be rejected on the basis of three groups of data. First, the differences between the expected phenotypic frequencies of offspring of high NTPH X low NTPH parental pairs (as calculated using the population ratio method based on the population distribution of NTPH activity in Figure 11) and the observed phenotypic frequencies are highly significant. Secondly, while this hypothesis predicts no high NTPH children of low NTPH X low NTPH parents, family T (Figure 17b) offers contrary evidence. Since only one child (T-III-l) had high NTPH activity, samples of blood from the parents and children were typed for HLA antigens to examine the possibility of non-parentage. Unfortunately, due to loss of lymphocyte viability during transport, only ABC and not HLA typing was possible for the blood sample of child (T-III-l). The results of the HLA and ABC typing shown in Table 10, provide no evidence for non— parentage of any of the children, but the lack of the critical HLA data for T-III-l renders this test inconclusive. The third body of evidence against Hypothesis I is the thermostability data presented in Section IV. C. 4. Because two classes of individuals can be distinguished by NTPH 123 thermostability in the low NTPH region, the low NTPH region cannot be represented by one genotype such as Hypothesis I suggests. For the purposes of using the population ratio method of analysis of family data, Hypothesis II which suggests that high NTPH activity is recessive and low NTPH is dominant, yields the same results as a codominant hypothesis where two genotypes can be discerned in the low NTPH region and one in the high NTPH region. This is detailed in Table 11. The thermostability data support this modification of Hypothesis II as does the comparison of the family study results with the expectations of the population ratio method of analysis, ignoring for a moment the two low NTPH offspring of a high X high parental pair. In addition, the pheno- typic frequencies of the NTPH specific activity regions (Figure 11) of 0 - 5 units, 5 - 25 units and >‘25 units, fit well the calculated Hardy—Weinberg equilibrium fre— quencies for a two allele system. However, when this codo- minant modification of Hypothesis II is proposed, it presents somewhat altered predictions for inheritance of NTPH levels in individual pedigrees. The addition of the 0 - 5 NTPH class, hereafter termed "very low NTPH', brings out dis- crepancies between this hypothesis and some of the pedigree information. Six families present evidence which conflicts with the codominant hypothesis presented in Table Ll. 124 TABLE 10. HLA and ABC Blood Typing of Family T Members Possible HLA ABO Type Haplotypes Father T-II-l A1 A2 A2 B27 Bw51 Mother T-II-Z A1 A1 Aw24 B8 B14 -- Cw2 Son T-III-l A1 Son T-III-Z Al A2 Aw24 B27 B14 -- Cw2 Son T-III-3 0 A2 Aw24 Bw51 B14 -- Cw2 TABLE 11. Codominant Hypothesis of Inheritance NTPH Variability NTPH specific activity (units/mg hemoglobin) 0-5 5-25 >25 Genotype N'N' NN' NN Population freq. from Fig. 11 .019 .153 .828 Hardy-Weinberg equilibrium freq..009 .173 .818 125 Four of these involve a very low NTPH individual which is incompatible with other family members. The first son (P-III-l) of family P (Figure 17a) and one of the daughters (Wa-II-S) of family Wa (Figure 18) are very low NTPH but in each case, one of the parents has high NTPH activity. Under the codominant hypothesis, the high NTPH parent has no alleles in common with the very low NTPH children. In families L (Figure 17a) and BE XFigure 18), there are parental pairs consisting of one high NTPH and one very low NTPH parent. According to the codominant hypothesis, all the offspring should be low NTPH, yet six of eight children have high NTPH activity. Additionally, in Family L, the very low NTPH individual (L-II-Z) and a high NTPH parent (L-I-3), the same situation as families P and Wa. The other two families that present evidence conflicting with the codominant hypothesis are H (Figure 17a) and Bo (Figure 18). Each family has two high NTPH parents, yet have at least one low NTPH child, an impossibility under the codominant hypothesis. The possibility of non-parentage and misclassification of subjects exists here, but with six families involved, the weight of evidence is against the codominant hypothesis. To permit a more complete analysis of these data, it was possible to perform thermostability studies on memebers of families H and L. The thermostability characteristics and NTPH activity of these families are shown in Figure 23. 126 Fomi ly L I IS a3 -O.606 Qt. l 2 .U ‘7l-— 4 -O.4|O 4.034 I III I '21? -O.434 -O.442 -O.442 -O.558 n.t. I 2 3 4 6 Fomi I y H I -4 -O.463 -o.453 I 2 II [315! ® 69 n. t. 6.528 n I. -o.595 I 2 3 4 Figure 23. NTPH Specific Activity and Thermostability of Two Families NTPH activity is inside the symbols expressed as units/mg hemoglobin. Thermostability slopes below the symbols were calculated as per Section IV. C. 4. and are expressed as log nmoles ITP cleaved per 10 minutes. n.t. = not tested. 127 Each individual exhibits thermostability characteristics of the NTPH activity range he or she belongs to as deter- mined in Section IV. C. 4. These data demonstrate that the NTPH specific activity of an individual as confirmed by thermostability characteristics provides an accurate assessment of an individual's phenotype group. Unfortunately, it was not possible to perform this additional analysis on all key families or family members of somewhat doubtful pheontypic status. From the discussion presented above, it is clear that neither simple one gene-two allele system of inheritance of NTPH variability conforms adequately to the pedigree and thermostability data. The following analysis of these data postulates that the variation of NTPH activity levels in human red blood cells can best be explained by the existence of three alleles at one NTPH structural gene locus. It has been suggested earlier in this thesis that the high NTPH region may be composed of more than one genotype. Figure 24 shows the distribution of the NTPH activity of the offspring of various combinations of parental NTPH activities. These data are drawn from Figures 17 and 18 and as before exclude the prOpositi of the families of Figure 18. The high NTPH region has been divided into two groups, 26 - 50 specific activity ufiits and > 50 specific activity units, based on indications of two modes Frequency (Number of Individuals) 128 4 ' i 1. Parental Type 6'25 X 26' 50 L,CC,ZO,Ca,Mc,Wa ,Pu 3 El 51 5 . II. Parental Type 6-25 X ) 50 q 0" J 7M'DO1 P7 T1w8v’ Y,H,]1,MM,Bu,Me,Pu,An III-I - IEI. Parental Type [7T 26-50 X 26-50 1 l J f1 F'I H,KK,Bo,So — ISZ. Parental Type ‘ 26-50 X >50 A,F,G,N,X,Z,BB,Li,Sa C)<3 to -b CDC3f0 -§C) 00 J5 O) a: "I ‘ Y. Parental Type 1 >50 X >50 B’C,E'I'K’Q’R’U' AA,DD,EE,FF,JJ, __ a . LL,NN . . —r- 7741 IO 20 30 4O 50 60 70 80 9O NTPH SPECIFIC ACTIVITY (units/mg hemoglobin) C) “) $><fi (D 1 Figure 24. NTPH Activity Distributions of Offspring of Various Parental Pairs In the five groups above, the NTPH specific activity ranges of the parental pairs are given and the families that comprise these groups are listed. The dotted line divides the "high" and "low" NTPH activity ranges. 129 of activity in Figures 11 and 20. If the high NTPH region were one genotype, one would expect that the offspring of any grouping of high NTPH parental pairs would mimic the population distribution in the high NTPH region. However, the distributions in Figure 24 show that this is not the case. A remarkable difference is seen between distributions I and II, which represent offspring of parental pairs composed of a low NTPH parent and a high NTPH parent of specific activity 26-50 or > 50. The mean high NTPH activity in distribution I is 49.1 while the mean high NTPH activity of distribution II is 66.3. A similar effect is seen in distributions III, IV, and V, which represent offspring of parental pairs consisting of two high NTPH parents in combination from the NTPH activity ranges of 26-50 and > 50. The mean high NTPH activity in these distributions increases as the activity of the parental pairs increase. The mean activities of distributions III, IV and‘V are 46.5, 53.6 and 63.0, respectively. These data strongly suggest the presence of heterogeneity within the high NTPH specific activity region. In relation to the arguments already presented concerning heterogeneity in the low NTPH region, the postulation of a system of in— heritance of NTPH activity controlled by multiple alleles at the structural gene loci for NTPH is required. 130 The simplest model that can be postulated in light of the rejection of both possible one gene-two allele models of inheritance of NTPH activity is one that involves three alleles at one NTPH structural gene locus. Three alleles at one locus provide six possible genotypes that must be adapted to the data presented. The model must provide for the existence of three phenotypic groups of thermostability in NTPH levels and also must account for the apparent heterogeneity in the high NTPH region of the population distribution. A proposed three allele system of the inheritance of NTPH activity is presented in Table 12. The three alleles N9, N' and N2 represent proposed null, low activity and high activity alleles. The NTPH activity ranges and the thermostability character- istics of each of the proposed genotypes are shown in Table 12. The proposed N'N2 genotype (NTPH specific activity 25-50) actually exhibits a slightly higher thermolability than the NZN2 genotype (NTPH specific activity > 50), the denaturation slopes (i S.E.) being -o.439 (e'o.oz3) and —o.423 (a 0.021) (data from Table 6). But the t-test of difference between two means reveals that the difference is not statistically significant ( (t26 - 1.725, 0.1>P>0.05). The family data support this proposal very well with only two possible exceptions. The following families offer examples of data that support the hypothesis that three alleles at the structural locus TABLE 12. 131 Specifié Activity A Model of the Inheritance of NTPH Specific Activity Range Thermostability units/mg hemoglobin Traits N0 Null N' Low N2 High Genotype NONo 0 -- NON' 1-7 Group 1 NON2 7-25 Group 2 N'N' 7-25 Group 2 WM2 26-50 Group 3 NZN2 >50 Group 3 Thermostability characteristics Group 1 Group 2 Group 3 slope = -0.968 (20.228) slope = -0.551 (20.038) slope = -0.433 (10.021) 132 for NTPH control the inheritance of NTPH activity. Families L (Figure 17a), W (Figure 17b), PP (Figure 17c) and Be (Figure 18) present parental pairs of very low NTPH, genotype NON', and high NTPH, genotype NZNZ. Of the very low NTPH parents, only the mother of family L (L-II-Z) and the father of Family W (W-I-l) have been examined for thermostability characteristics and in both persons the results indicated that they did indeed belong in the NON' genotype. In these families all the offspring should be NON2 (NTPH specific activity 7-25) or N'N2 (NTPH specific activity 26-50), and the results bear out these predictions for each child. Families P (Figure 17a) and CC (Figure 17c) represent parental pairs of genotype N°A2 and N'NZ. The mother of Family P (P-II-Z) represents an example of the overlapping ranges of the genotypes N'N2 and NZN2 since the first son (P-IV-B) by thermostability criteria is of genotype NON'. The mother would have to possess an N' allele under the proposed hypothesis. Offspring of the parental pairs of Families P and CC could have four possible NTPH genotypes: NON', N0,N2, N'Nz, or N2N2. Indeed, the five children in these two families exhibit these four genotypes among them. That the N'N' genotype is in the NTPH specific activity region of 7-25 and is thermolabile is shown by Families H((Figure 17a) and Bo (Figure 18). Each of these families exhibits each of these genotypes. Additionally, the thermostability 133 characteristics of the two postulated N'N' children in Family H (H-II-Z add H-II-4) were in accordance with the 7-25 NTPH specific activity group as a whole (See Figure 23). Family T (Figure 17b) is the only example of a family with both parents' NTPH specific activity in the 7-25 range. It must be postulated that their genotypes are both NON2 and not N'N' since the three offspring express the three genotypes expected from two N°N2 parents. This family presents evidence for the NONO genotype expressing no NTPH activity in red blood cells. Repeat samples of the child with 0 NTPH activity consistently presented this result. The heterogeneity of the high NTPH region is exemplified in Family Z (Figure 17b). One parent is in the activity range of genotype N'N2 and the other in the range of genotype N2N2. Of the six children, three have NTPH levels like the N'N2 parent and three have NTPH levels like that of the NZN2 parent. In this family the NTPH specific activities of these two genotypes are widely separated. A final example is that of Family II (Figure 17c) where the parental genotypes are NZN2 and possibly N'N'. The son has intermediate NTPH activity and could represent the genotype N'NZ. There are only two individuals in all of the pedigrees whose NTPH specific activities are iincompatible with the genotype assignments of other family members. These both involve families obtained from J.F. Henderson's laboratory. 134 In family Ke (Figure 18) the parental genotypes would appear to be NON' and NZNZ, yet one child (Ke-II-l) is apparently NON', an impossibility unless overlapping of genotypes of either parent or the child is causing mis- classification of these family members. Blood samples were not available for the testing of thermostability. The second incompatibility is the presence in Family Wa (Figure 18) of a child of 0 NTPH activity (Wa-II-S), genotype N°N°. The genotype of the mother (Wa-I-Z) would be N'NZ, incompatible with that of the child. A possible explanation here could be in analysis of NTPH activity in this subject. The end point NTPH assay system used in these studies may not always distinguish very low activity from no activity, due to small errors in reagent blank correction, such that this apparent N°N° individual may be a mis- classified NON' genotype, which would then be compatible with the parents. 135 V. Discussion A. Physical Studies of NTPH Various studies of physical parameters of human red cell NTPH have been presented in this dissertation which further characterize the enzyme. Isoelectric focusing of NTPH from red cell lysates and from partially-purified pre- parations (SOD-fold purification) show a pI of about 4,7 for NTPH (Figures 1 and 2). This finding is in close agree- ment with the pI of 4.3 - 4.5 for rabbit NTPH ascertained by Chern (75). A purification scheme is also presented here which results in a nearly homogenous preparation of NTPH. Figure 3 shows the protein stained banding pattern of a native polyacrylamide gel of a preparation of purified NTPH. Two major protein bands are present which constitute about 90% of the total protein on the gel. NTPH activity staining corresponds to one of the bands. However, when these two bands are cut out of stained gels, the protein eluted and applied to SDS-polyacrylamide gels, the banding patterns obtained for each of their proteins are identical (Figure 7). Each shows two bands on SDS gels corresponding to molecular weights of 26,000 and 40,000 daltons. More recent puri- fication procedures using dithiothreitol instead of 2- mercaptoethanol have resulted in elimination of the 26,000 dalton polypeptide as a contaminant. Thus, a subunit molecular weight of 40,000 is proposed for NTPH. This 136 figure is slightly larger than the 34,000 - 35,000 molecular weight determined by sucrose density gradient centrifugation (Figures 5 and 6). The lower molecular weight obtained by the latter technique may be due to the molecular shape of NTPH. The gel filtration evidence (Figure 4) that NTPH is larger than hemoglobin (a 65,000 daltons) suggests that native NTPH is a dimer composed of identical 40,000 dalton subunits. The thiol requirements of human red cell NTPH have been re-examined in light of a report by Vanderheiden which stated that the enzyme had no requirement for thiol compounds (7). This observation was contrary to the experience of this laboratory. Table 1 shows that pre-incubation of NTPH of lysates and highly purified preparations with varying con- centrations of the thiol reagents N-ethylmaleimide (NEM) iodoacetamide (IAM) and p-hydroxymercuribenzoate (pHMB) revealed marked sensitivity of NTPH to each of these reagents. Incubation with 50 mM NEM and 10 mM pHMB produced total inhibition of NTPH activity while 70% inhibition of activity was achieved with 50 mM IAA. In addition, the ability of the thiol compounds 2- mercaptoethanol, glutathione and dithiothreitol (DTT) to stabilize NTPH activity during prolonged dialysis was examined. The data of Table 2 show that NTPH activity was stabilized much more effectively by DTT than by the other thiol compounds. Also, dialyzed NTPH preparations showed 137 significantly higher NTPH activity when analyzed in the presence of 2 mM DTT than when no DTT was added to a reaction mixture. Figures 9 and 10 show that in a 20 minute incubation at 37°C, NTPH activity is linear in the presence of DTT, but that the enzyme is inactivated in an assay lacking DTT. Partial reactivation of inactive NTPH with DTT could be demonstrated by addition of DTT to an NTPH preparation which had been incubated in the absence of DTT until complete in— activation had occurred. These data support the conclusion that human red cell NTPH has an absolute requirement for the presence of a reduced thiol compound and that DTT is the thiol compound most effective in stabilizing NTPH activity. B. Analysis of the Genetic Variation of Human Red Cell NTPH Activity Early work in this laboratory on the purification and characterization of rabbit red cell NTPH (1) also involved examination of the enzyme in human erythrocytes. It was observed that there existed a vast degree of variation in NTPH activity among individuals, but that in any one individual, this activity was constant over several years sampling. Vanderheiden, in an indirect manner, provided the first evidence that such variation was inherited (3). He found two siblings who exhibited elevated levels of ITP in hemolysates and proposed that this condition was due to an inherited trait. He followed this with a population survey and a few family studies of an "ITPase" activity (4). 138 Erythrocytes of subjects with deficient "ITPase" activity were able to synthesize high levels of [14C] ITP when incubated with [14C] inosine. It was concluded from the population survey data and from the pedigrees presented that "ITPase" activity could be classified into three phenotypes and that two dodominant alleles, one for high activity and one for low, controlled inheritance of "ITPase" activity levels. Vanderheiden's "ITPase" activity, as measured by production of inorganic phosphate, was actually the coupled reaction of NTPH and endogenous inorganic pyrophosphatase. In addition, the methods employed in choice of subjects and assay conditions for "ITPase" activity cause serious reservations to be entertained concerning the conclusions reached. The samples for the population survey and for the family studies were drawn from a population of mixed racial origins, a situation which can prove inaccurate, if racial differences exist, as exemplified by the racial differences in galactokinase activity (57). Also, many of the pedigrees are lacking assays of one or both parents of a particular family and some of the proposed genotypes of family members are incompatible with the genetic hypothesis presented. The assay method for "ITPase" provided sub- optimal conditions for the accurate assessment of NTPH activity. It was of interest, therefore, to examine the variability of NTPH activity under optimal conditions in order to provide confirmation or rejection of the "ITPase" hypothesis. 139 A population survey of the NTPH activity distribution in the red blood cells of a random Caucasian population was undertaken which clearly demonstrated the presence of at least two modes of activity -- high and low NTPH specific activity regions representing 82.8% and 17.2% of the population, respectively (Figure 11). This distribution, which corresponds fairly well to the "ITPase" data presented by Vanderheiden, conflicts with his later data which exhibits a unimodal population (19). No explanation has been presented by Vanderheiden for the discrepancy in his two sets of data. The data presented in Figure 11 provided a basis for the biochemical and genetic examinations of the high and low NTPH phenotypes. Previous data of this laboratory have demonstrated the presence of NTPH in 11 tissues of the rabbit (2). Likewise, NTPH activity has recently been demonstrated in human lympho- cytes, granulocytes and platelets (6). The variation in the NTPH levels expressed in red cells is reflected in these cells as well, thereby indicating that the biochemical basis of the observed diversity is not limited to red cell development of metabolism. The quantitative variation of NTPH activity may be caused by a number of molecular factors which fall into three classifications. First, an altered NTPH protein may be present which has concomitantly altered kinetic parameters. Second, the NTPH protein in individuals may be identical but may differ in the actual amount present 140 due to alterations in synthetic or degradative rates. Third, the variation in NTPH activity may be due to the presence of intracellular activators or inhibitors in different individuals. Some of these possibilities have previously been examined. V. Verhoef of this laboratory searched for the presence of intracellular effectors by mixing hemolysates of individuals of different NTPH activities. The Strictly additive nature of the results provided no evidence for such activators or inhibitors (81). He also examined the Km for ITP in hemolysates of several high and low NTPH subjects and found no pattern of difference in the two groups (5). The search for biochemical differences in the NTPH of individuals with varying red cell specific activity has included examination of electrophoretic patterns for iso- zymes. Wang and Morris (2) found no evidence for isozymes in rabbit red cell, rabbit liver, or human red cell NTPH preparations. Harris and Hopkinson (15) report the existence of a rare electrophoretic variant, but no common variants that correspond to the quantitative variation of activity. Further experimentation which examines possible molecular bases of NTPH variation is presented in this dissertation. NTPH isozymes were not found by isoelectric focusing in polyacrylamide gels, a technique of greater resolution than polyacrylamide gel electrophoresis. The utilization of the alternative substrates dITP and XTP was 141 examined in hemolysates of high and low NTPH individuals. Each subject utilized the alternate substrates in a manner corresponding to the extent of ITP hydrolysis by that hemolysate (See Table 4). The in inQ_ stability of the NTPH enzyme activity in subjects covering a wide range of specific activities was examined by analysis of loss of NTPH activity during the lifetime of the red cell. This was accomplished by separating erythrocytes of an individual into populations of cells of different mean age by their density, which is age-related. The data of Table 5 indicate that loss of NTPH activity with red cell age is similar over the range of NTPH specific activity values tested, such that increased or decreased in yiyg stability of NTPH cannot account for the quantitative variation in NTPH activity observed. The analysis of thermostability characteristics of NTPH in individuals of diverse activity levels has provided the first evidence of a qualitative difference in NTPH which is associated with the quantitative variation. Three distinct statistically different phenotypes have been distinguished in hemolySates heated at 65°C. The high NTPH individuals (> 25 units specific activity) exhibit the most heat stable enzyme, while the low NTPH range is now divided into two phenotypes -- one with specific activity of 0-5 and the most heat labile enzyme, and another with specific activity of 5-25 which exhibit intermediate thermostability characteristics (See Tables 6 and 7 and Figures 14-16). 142 The mode of inheritance of the quantitative variation of NTPH activity has been analyzed using the data of the pOpulation distribution of NTPH activity, the thermostability characteristics, and the study of NTPH activity in family members. Analyses of NTPH activity in 57 families have been carried out. There is no evidence of sex-linked transmission of NTPH activity as the distributions of activity in males and females are similar and several families exhibit father to son transmission of low NTPH activity. That inheritance of NTPH activity is autosomal is also supported by two additional pieces of information: 1) the structural locus for NTPH has somewhat tentatively been assigned to chromo- some 20 (16,17) and 2) the altered thermostability character- istics of the very low and low NTPH groups are evidence for an allele of the structural gene which, if indeed located on chromosome 20, indicates an tutosomal trait. Two possible one gene-two allele systems of inheritance were examined and both must be rejected on the basis of the available data. The first hypothesis, that low NTPH is one homozygous genotype and high NTPH is composed of hertero- zygotes and homozygotes for a high NTPH allele, is rejected by three bodies of evidence. First, the thermostability data indicate the existence of two phenotypic groups within the low NTPH region. Second, the observed phenotypic frequencies of offspring of high X low NTPH parental pairs are highly significantly different from the expected frequencies calculated using the population ratio method of Li (80) 143 (See Tables 8 and 9). Third, family T (Figure 17b) provides pedigree data that contradict this hypothesis. The second hypothesis, that the low NTPH region is composed of two geno- types and that the high NTPH region is one group, is compatible with thermostability data and expected frequencies of the population rationmethod, but is contradicted by data from six families, the H, L, P, Be, Wa, and Bo families (Figures 17b and 18) as well as evidence from offspring groups of high X high NTPH parental pairs that suggest heterogeneity of the high NTPH region (See Figure 24). The data from the families above indicate the presence of at least four geno- types. It is proposed, then, that the quantitative variation of NTPH activity is genetically controlled by multiple alleles at a structural locus for NTPH. I Specifically, it is proposed that the genetic variation of NTPH activity is controlled by three alleles at one of the structural loci for NTPH as described in Table 12. The three alleles are NO, N' and N2 yielding six genotypes -- N°N°, NON', NONZ, N'N', N'Nz, and N2N2. The NONO genotype is uncommon, 1%, and is postulated to have no detectable NTPH activity. The NON' genotype has NTPH specific activity between 1 en.2 and approximately 7 units/mg hemoglobin. This genotype also expresses a much faster rate of thermal inactivation than the other genotypes. The NTPH specific region between 7 and 25 units/mg hemoglobin is proposed to be comprised of the NON2 and N'N' genotypes, both of which exhibit increased 144 thermolability significantly different than those individuals above specific activity of 25, but not nearly the extent of thermolability of NONl individuals. The N0 allele, then, is thought to synthesize a heat labile produce while the N' allele product is somewhat intermediate to the N0 and N2 alleles in thermostability. The N'N2 genotype represents NTPH specific activity of 26-50 units/mg hemoglobin. The mean thermostability of this genotype is slightly less than that of the remainder of the high NTPH group (N2N2) but the difference is not statistically significant. The remaining genotype, N2N2, expresses NTPH specific activity above 50 units/mg hemoglobin. It is anticipated from the population distributions of Figures 11 and 20 that considerable overlapping in the NTPH specific activities of individuals with N'N2 and NZN2 genotypes exists. This three allele hypothesis is the simplest method of inheritance that can account for the data presented. The evidence for this hypothesis can be summarized as follows: 1) thermostability and family studies provide evidence of at least four distinct phenotypes, two in the low NTPH region (distinguished by thermostability studies) and two in the high NTPH region (distinguished by family studies, particularly families L, W and Wa, and by the distribution of offspring of various high X high NTPH parental pairs as shown in Figure 24), 2) simple one gene- 145 two allele hypotheses can be ruled out by their incompatibility with this data, and 3) there are only two individuals in the family study whose NTPH specific activity values are not compatible with this hypothesis. That N0, N', and N2 are alleles of the structural locus of NTPH is supported by the observed differences in thermostability associated with the No and N' alleles compared to the "normal" N2 allele. Such a difference suggests an alteration of the gene product produced by the No and N' alleles rather than a regulatory gene phenomenon. Figure 25 shows the eight families that were inconsistent with the one gene-two allele hypothesis of NTPH activity variation, plus family Ke. Genotypes have been assigned under the proposed three allele system of NTPH inheritance. One member of Family Wa, marked by the arros, is incompatible with the genotypes assigned to the parents. As noted earlier, this may represent a small error in the assay value obtained, since very low and zero NTPH represent very similar Optical density readings in our assay system. The other individual that does not fit genotypes of other family members is that marked by the arrow in Family Ke. Misassignment of genotypes due to overlapping NTPH activity in either parent or the child in question could be the cause of this in- compatibility. The mother of Family P is thought to re— present an eXample of the overlap of genotypes N'N2 and NZN2 since under the proposed pattern of inheritance, she is an obligate heterozygote. 146 a—a N"'Nz N°N‘ I-l 29115] I E} N°N" ' N°N‘ N‘N' E, [a _ N‘N‘ N°N‘ N°N° Ll. E——® EQ. N'N'I N'N‘ 69 e is a N'N’ N'N‘ N'N' N'N‘ N'N’ Figure 25. Proposed Genotypes of Members of Nine Families The NTPH specific activity appears in the circles and squares and the genotypes underneath. Thermostability characteristics, if tested are signified l, 2, or 3 as per Table 12. Arrows indicate individuals whose genOtype is incompatible with the rest of the family. 147 I. IBI—-® .2 —® . NgN N‘ N’ N°N‘ N' N" are 125.]— E NgN‘ N‘l’N' NI’N' N' N' a 53 E (55 E N'N' N'N‘ N'N‘ N°N' N'N‘ 3 3 3 2 w a... e are @ I 6}) O I15 a N'N‘ . N°N’ N°N' N' N' N‘ N' M EF-® 5: EB——@ N°N‘ N'N‘ N°NI Nst asap 3 IE?) N'N‘ N°N' N°N‘ N°N‘ N:N N' N'N' N°N‘ NiN' Figure 25(Cont'd) 148 Overall, the family data support the hypothesis of three alleles at one locus very well. These data represent evidence against the codominant two-allele theory presented by Vanderheiden (4) to explain "ITPase" (NTPH) variation in activity. The possibility exists that there are more than three alleles determining NTPH activity. In particular, there may be another high NTPH allele suggested by the lack of fit to a normal distribution of the distribution of NTPH activity above 50 units/mg hemoglobin in Figure 20. It is proposed here that a one gene—three allele system of inheritance of NTPH activity as outlined in Table 12 represents a simple and accurate assessment of the genetic and biochemical data obtained to this time. C. The Metabolic Role of NTPH and Consequences of Its Genetic Variation The physiological role of NTPH in the cell is a matter of considerable interest and speculation at this point. While most hypotheses of the metabolic significance of NTPH concern the effects of its most readily utilized substrates, ITP and dITP, Hershko et_al. (12) have suggested that the actual physiological substrate is GTP. This nucleotide is hydrolyzed in rabbit and human red cells at rates 10% and < 2% that of ITP, respectively (1, 7, 8). There is considerable evidence, however, that ITP is a physiological substrate for NTPH. Vanderheiden (3, 4) found elevated ITP levels in red blood cells of ‘ 0 149 individuals and attributed this observation to be the result of "ITPase" (NTPH t endogenous inorganic pyrophosphatase) deficiency. He also studied accumulation of [14C] ITP in red blood cells incubated with [14C] inosine and noted that [14C] ITP levels were highest in "ITPase" deficient individuals. Collaborative studies involving this laboratory examined the relationship between NTPH activity and accumulation of [14C] ITP in erythrocytes incubated with [14C] hypozanthine (5). In general, there was an inverse relationship between [14C] ITP accumulation and NTPH activity that followed closely a theoretical relationship for a substrate and its enzyme as predicted by Michealis-Menten kinetics. Additionally, ITP has been observed directly by high pressure liquid chromatography of a PCA extract of blood from an individual with very low NTPH specific activity (3 units/mg hemoglobin) (81). The ITP concentration in this sample was calculated to be approximately 5 pM while no ITP was observed by the same methods in erythrocytes of individuals with higher NTPH specific activities of 15 and 35 units/mg hemoglobin. Additionally, the amount of GTP present was not observed to vary among the individuals examined. Although several investigators have shown that erythrocytes from stored or fresh blood can accumulate ITP in high concentrations (even reaching 1.4 mM, high than the w ’_ ~ " —_'- "_'—'"'__"-' 150 physiological concentration of ATP) (82 - 90), there is yet no known physiological role of ITP in the cell. Vanderheiden has suggested that ITP plays a role in an "ITP : IMP cycle" which is proposed to be a regulating mechanism for intra- cellular levels of ATP (14). In this scheme, ITP is formed from IMP by transfer of phosphate groups from ATP in a two step proeedure. ITP is then hydrolyzed to IMP by NTPH. The net reaction is essentially an ATPase-like reaction which yields AMP and 2 Pi. No direct evidence is presented for this proposal. The consequences of high intracellular levels of ITP due to low activity of NTPH are not known. Fraser et 31. (18) examined the accumulation of [14C] ITP from [14C] hypoxanthine in erythrocytes. Erythrocytes from 5% of a normal pOpulation accumulated "high" levels of [14C] ITP while the incidence of this characteristic in a mentally retarded population was 16%. NTPH levels were not tested. The data of Soder gt 31. (5) suggest that individuals exhibiting "high" ITP accumulation were probably low in NTPH activity. Vanderheiden has presented data concerning a higher proportion of NTPH—deficient individuals among paranoid shhiZOphrenics than in the normal population (19). He has proposed that high ITP levels inhibit the activity of glutamic acid carboxylase, an enzyme found to be reduced in activity in the brains of sehizophrenics and psychotics (20). The data presented in Table 3 of this dissertation, which 151 examined the red cells of patients with schizophrenia and paranoid schizophrenia provides no evidence of striking differences between the NTPH activities of this sample and the distribution found in the normal population (Figure 11). A different consequence of high ITP levels was suggested by Wang and Morris (2) when they proposed that the metabolic role of NTPH may be to prevent incorporation of ITP into RNA and dITP into DNA by keeping levels of these nucleotides low in the cells. ITP can be used as a substrate in the formation of polynucleotides catalyzed by RNA polymerase from Azobacter vinlandii (91). Similarly, in E; coli, DNA polymerase incorporates dIMP into DNA in place of dGMP when dTTP is supplied as substrate (92). In mammalian systems, dITP can serve as substrate for calf thymus DNA polymerase-catalyzed polymer formation (93). A situation analogous to the role of NTPH as proposed by Wang and Morris occurs in E; ggli_mutants deficient in dUTPase activity. This enzyme catalyzes the pyrophosphory- lytic cleavage of dUTP similar to the reaction catalyzed by NTPH using ITP and dITP (94). Mutants of §;_coli deficient in dUTPase accumulate short Okazaki fragments in replications of DNA (95). The presence of the short fragments has been explained as the misincorporation of dUTP into DNA as a result of the increased availability of dUTP in these mutants, followed by its excision and repair. Short Okazaki fragments have also been observed 152 in the replication of polyoma DNA when dUTP replaced dTTP in the in yit£g_synthetic system (96). The first step in excision of the dUMP residue is thought to be catalyzed by uracil-DNA glycosylase, which cleaves uracil from the deoxyribese of the DNA backbone (97). This apyrimidinic site is then thought to be cleaved by an endonuclease and then repaired by an exonuclease-initiated process. Double ‘mutants of uracil-DNA glycosylase and dUTPase do not accumulate short Okazaki fragments (95). Uracil-DNA glycosylase is active in the repair of deaminated cytosine residues (uracil) and presumably will operate in the same manner of uracil residues that have been incorporated into the DNA directly (98). Thomase EE.E£- (99) have recently reported similarly short Okazaki fragments in an §;_ggli in yit£g_DNA:synthesizing system when dITP is added. Analagous to the dUTP studies, a hypoxanthine-DNA glycosylase has beenffound in E; 921; which catalyzes the excision of hyposanthine from the DNA backbone (100). This enzyme activity is clearly distinct from that of the earlier observed 3-methyladenine-DNA glycosylase (101) or the uracil-DNA_glycosylase (97). There is recent evidence of a "purine insertase" activity which incorporates purine, but not pyrimidine, bases specifically into apurinic sites on the DNA (102). This enzyme is template specific, utilizing guanine, but not adenine in depurinated poly (dG-dC) and incorporating 153 adenine, but not guanine in depurinated poly (dA-dT). This activity represents an alternative pathwayy for DNA repair, substituting for excision processes. Thus E; coli can both incorporate and excise dITP from DNA. If these mechanisms for excision of dIMP from DNA can be demonstrated in mammalian cells, one can postulate that low or non- existent levels of NTPH in the cell with the concomitantly higher ITP and dITP levéls could overwhelm these repair processes. Our data indicate that approximately 1% of the Caucasian population have red blood cells in which NTPH is undetectable by our analytical procedure. Since red blood cells do not synthesize DNA or RNA it has been of interest to determine whether other cell types express similar deficiency of activity. An individual with very low red cell NTPH activity had granylocyte and lymphocyte NTPH levels 5-fold and 30-fold, respectively, that of the red cells when compared on the basis of NTPH activity per cell (6). While ITP was detectable in the red cells of this subject, the other cell types were not examined for ITP levels. Such tissue differences in NTPH activity per cell may prevent ITP of dITP accumulation in those cells capable of synthesizing DNA and RNA. It is noteworthy that tissue studies in the rabbit reveal that erythrocytes have the lowest per cell NTPH activity of all tissues studied (2). Unfortunately, blood samples from the un- common individuals with no detectable erythrocyte NTPH 154 activity were not available for examination of NTPH in granulocytes and lymphOcytes as above. If these cell types are also totally deficient in NTPH activity, the possibility of high intracellular concentrations of ITP and dITP exists for these individuals and physiological effects of this genetic variability of NTPH activity may yet be delineated in a small portion of the human population. LIST OF REFERENCES 10. 11. 12. 13. 14: 15. LIST OF REFERENCES Chern, J., MacDonald, A.B., and Morris, A.J. (1969) J. Biol. Chem. 244, 5489-5495. Wang, J.K., and Morris, A.J. (1974) Arch. Biochem. Vanderheiden, B.S. (1965) Biochem. Biophys. Commun. 21, 265-270. Vanderheiden, B.S. (1969) Biochem. Genet. 3, 289-297. Soder, C., Henderson, J.F., Zambor, G., McCoy, E., Verhoef, V., and Morris, A.J. (1976) Can. J. Biochem. a, 843-847. Verhoef, V., Fuller, S.A.,aand Morris, A.&. (1980) Biochem. Genet. 18, 235-246. Vanderheiden, B.S. (1979) J. Cell, Physiol. 28, 41-48. Morris, A.J. (1978) in Method in Enzymology (Hoffee, P.P. and Jones, M.E., eds.), Vol. 51, pp. 275-285, Academic Press, Inc., New York. Liakopoulou, A., and Alivisatos, S.G.A. (1964) Biobhim. Biophys. Acta 89, 158-161. Hershko, A., Radin, A., Shoshani, T., and Maher, J. (1967) Biochim. Biophys. Acta 149, 59-73. Pynes, G.D., and Younathan, E.S. (1967) J. Biol. Chem. 242, 2119-2123. Hershko, A., Jabotinsky, K., and Mager, J. (1969) Isr. J. Med. Sci. 5, 991-997. Vanderheiden, B.S. (1970) Biochim. Biophys. Acta 215, 555-558. Vanderheiden, B.S. (1975) J. Celll Physiol.§6, 167-175. Harris, H., and Hopkinson, D.A. (1976) Handbook of Enzyme Electrophoresis in Human Genetics, North-Holland, Amsterdam. 155 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 156 Meera-Khan, P., Pearson, P.L., Wignen, L.L.L., Doopert, B.A., Westeveld, A., and Bootsma, D. (1976) Cytogenet. and Cell Genet. 16, 420-421. Hopkinson, D.A., Povey, 8., Solomon, E., Bobrow, M., and Gormley, I.P. (1976) Cytogenet. Cell Genet. 16, 159-160. Fraser, J.H., Meyers, H., Henderson, J.F., Brox, L.W., and McCoy, E.E. (1975) Clin. Biochem. 8, 355-364. Vanderheiden, B.S., and Zarane-Moyano, C. (1976) Biol. Psychiatry 11, 755-765. Vanderheiden, B.S. (1979) Biochem. Med. 21, 22-32. Bird, E.D., Barnes, J., Iversen, L.L., Sparkes, E.G., MacKay, A.V.P., and Shepherd, M. (1977) Lancet 2, 1157. Kelley, W.M., Rosenbloom, F.M., Henderson, J.F., and Seegmiller, J.E. (1967) Proc. Natl. Acad. Sci. U.S.A. 51, 1835. Okada, S. and O'Brien, J.S. (1969) Science 165, 698. Stanbury, J.B., Wyngaarden, J.B., and Fredrickson, D.S. (1978) The Metabolic Basis of Inherited Disease, 4th Edition, p. 14, McGraw Hill, New York. Weinshilboum, R. and Raymond, F.S. (1977) Am. J. Hum. Genet. 29, 125-135. Harris, H. (1975) Human Biochemical Genetics, pp. 147-186, North Holland, Amsterdam. Ingram, V.M. (1959) Biochim. Biophys. Acta 36, 406-411. Chang, J.C., and Kan, Y.W. (1979) Proc. Natl. Acad. SCi. U.S.A. _7_§-, 2886-2889. Temple, G.F., Chang, J.C., and Kan, Y.W. (1977) Proc. Natl. Acad. Sci. U.S.A. 14, 3047-3051. Itano“ H.A. (1957) Adv. Prot. Chem. l_2, 216. Boyer, S.H., Hathaway, P., and Garrick, M.D. (1964) Cold Spring Harbor Symp. Quant. Biol. 29, 333. Paigen, K. (1971) in Enzyme Synthesis and Degradation in Mammalian Systems (Rechcigl, M., ed.) p. 1, University Park Press, Baltimore. EcKusick, V. (1978) Meddelian Inheritance in Man, 5th Edition, John Hopkins, Baltimore. Seegmiller, J.B., Rosenbloom, F.M., and Kelley, W.N. (1967) Science, 155, 1682. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 157 Rubin, C.S., Dancia, J., Yip, L.C., Nowinski, R.C., and Balin, M.E. (1971) Proc. Natl. Acad. Sci. U.S.A. 65, 1461-1464. Arnold, W.J., Meade, J.C.,aand Kelley, W.N. (1972) Upchurch, K.S., Leyoa, A., Arnold, W.J., Holmes, E.W., and Kelley, W.N. (1975) Proc. Natl. Acad. Sci. U.S.A. 12, 4142-4146. Ghangas, G.S., and Milman, G. (1975) Proc. Natl. Acad. Sci. U.S.A. 12, 4147-4150. McDonald, J.A., and Kelley, W.N. (1971) Science 171, 689-691. Kelley, W.N., Greene, M.L., Rosenbloom, F.M., Henderson, J.F., and Seegmiller, J.E. (1969) Ann. Intl. Med. 19, 1556206. Carson, P.E., Flanagan, C.I., Ickes, C.E., and Alving, A.S. (1956) Science 124, 484. Beutler, E. (1978) in The Metabolic Basis of Inherited Disease (Stanburyy, J.B., Wyngaarden, J.B., and Fredrickson, D.S., eds.), pp. 1430-1451, McGraw-Hill, New York. Piomelli, S., Gorash,L.M., Davenport, D.D., Miraglia, J., and Amorosi, E.L. (1968) J. Clin. Invest. 41, 940-948. Morelli, A., Benatti, U., Goetani, G.F., and DeFlora, A. (1978) Proc. Natl. Acad. Sci. U.S.A. 15, 1979-1983. Dern, R.J. (1966) J. Lab. Clin. Med. 68, 560-565. Dern, R.J., McCurdy, P.R., and Yoshida, A. (1969) J. Lab. Clin. Mad. ‘13—, 283-2900 Yoshida, A. (1970) J. Mol. Biol. 32, 483-490. Valentine, W.N., Tanaka, K.R., and Miwa, S. (1962) Trans. Assoc. Am. Physicians 14, 100-110. Valentine, W.N., and Tanaka,K.R. (1978) in The metabolic Basis of Inherited Disease (Stafibury, J.B., wyngaarden, J.B., and Fredrickson, D.S., eds.), pp. 1410-1429, McGraw-Hill, New York. Black, J.A., Rittenberg, M.B., Bigley, R.H., and Koler, R.D. (1979) Am. J. Hum. Genet. 31, 300-310. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 158 Adachi, K., Ghory, P.K., Asakura, T., and Schwartz, E. (1977) Proc. Natl. Acad. Sci. U.S.A. 14, 501-504 Shaw, C.R. (1965) Science 149, 936. Spencer, N., Hopkinson, D.A., and Harris, H. (1964) Nature 201, 299. Luffman, J.E., and Harris, H. (1967) Ann. Hum. Genet. .14, 387. Lewis, W.H.P. (1973) Ann. Hum. Genet. 14, 267-271. Scanlon, P.D., Raymond, F.A., and Weinshilboum, R.M. (1979) Science 203, 63-65. Tedesco, T.A., Miller, K.L., Rawnsley, B.E., Menutli, M.T., Spielman, R.S., and Mellman, W.J. (1975) Am. J. Hum. Genet. 11, 737-747. Spielman, R.S., Harris, H., Mellman, W.J., and Gershowitz, H. (1978) Am. J. Hum. Genet. 14, 237—248. Hansen, J. (1977) Anal. Biochem. 14, 37. Morton, N.E. (1962) in Methodology in Human Genetics (Burdette, W.J., ed.YT pp. 17-52, Holden-Day, San Francisco. Rathbun, W.B., and Betlach, M.V. (1969) Anal. Biochem. 14, 436-445. Austin, J.H., and Drabkin, D.K. (1935) J. Biol. Chem. 112, 67-88. Lowry, OHH., Rosenbrough, N.J, Farr, A.L., and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. Martin, R.G., and Ames, B.N. (1961) J. Biol. Chem. 236, 1372-1379. Kunitz, M. (1950) J. Gen. Physiol. 11, 349 and 363. Chiancone, E., Vecchini, P., Forlani, L., Antonini, E., and Wyman, J. (1966) Biochim. Biophys. Acta 127, 549. Lindberg, U. (1967) Biochem. 4, 335. Gabriel, 0. (1971) 14 Methods in Enzymology (Jakoly, W.B., ed.), Vol. 22, pp. 565-578, Academic Press, New York. Weber, K., Pringle, J.R., and Osborn, M. (1972) 14 Methods in Enzymology (Hirs, C.H.W., and Timashoff, S.N., eds.), XXVI, pp. 3-27, Academic Press, New York. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 159 Blakesley, R.W., and Boezi, J.A. (1977) Anal. Biochem. 41, 580. Murphy, J.R. (1973) J. Lab. Clin. Med. 41, 334-341. Cohen, N.S., Ekholm, J.B., Luthia, M.G., and Hanahan, D.J. (1976) Biochim. Biophys. Acta 419, 229-242. Radola, B.J. (1974) Biochim. Biophys. Acta 386, 181-193. Radola, B.J. (1976) 14 Isoelectric Focusing (Catsimpoolas, ed.), pp. 119-171, Academic Press, New York. Chem, C.J. (1970) thesis: "A Nucleoside Triphosphate Pyrophosphohydrolase From Red Cells of the Rabbit", Michigan State University, E. Lansing. Winkrhalter, K.H., and Huchus, E.R. (1964) J. Biol. Chem. 239, 2699-2705. Croxton, F.E. (1953) Elementary Statistics with Applications in Medicine, Prentice-Hall, New York. Sokal, R.R., and Rohlf, F.J. (1969) Biometry, pp. 220-221, W. H. Freeman, San Francisco. Rohlf, F.J., and Sokal, R.R. (1969) Statistical Tables, W. H. Freeman, San Francisco. Li, C.C. (1961) Human Genetics, pp. 31-35, McGraw-Hill, New York. Verhoef, V.J. (1978) Thesis: "The Role of Nucleoside Triphosphate Pyrophosphohydrolase, A Genetically Variable Enzyme, in Inosine Triphosphate Metabolism in Human Erythrocytes", pp. 56-58, Michigan State University, E. Lansing. Blair, D.G.R., and Dommasch, M. (1969) Transfusion 1, 198-202. Zachara, B. (1974) J. Biochem. (Tokyo) 14, 891-895. Zachara, B, and Lewandowski, J. (1974) Biochim. Biophys. Acta 253, 253-259. Zachara, B. (1975) J. Lab. Clin. Med. 11, 436-444. Zachara, B. (1975) Vox. Sang. 14, 453-455. Agarwal, R.P., Crabtree, G.W., Parks, R.R., Nelson, J.A., Keightley, R., Parkman, R., Rosen, F.S., Stern, R.C., and Polmar, S.H. (1976) J. Clin. Invest. 11, 1025-1035. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 160 Helson, D.J., Bugge, C., and Krasny, H.C. (1976) Adv. Exp. Med. Biol. 76A, 121-128. Henderson, J.F., Zombor, G., Fraser, J.H., McCoy, E.E., Verhoef, V., and Morris, A.J. (1977) Can.J. Biochem. 11, 359-364. Vanderheiden, B.S. (1979) J. Cell Physiol. 11, 287-302. Krakow, J.S., and Karstadt, M. (1967) Proc. Natl. Acad. Sci. U.S.A. 14, 2094-2101. Bessman, M.J., Lehman, I.R., Adler, J., Zimmerman, S.D., Simms, E.S., and Kornberg, A. (1958) Proc. Natl. Acad. Sci. U.S.A. 44, 623-640. Bollum, F.J. (1966) 14 Proceedings, Nucleic Acid Res. (Cautoni, G.L., and Davis, D.R., eds.) pp. 577-583, Harper and Row, New York. Bertani, L.E., Haggmark, A., and Reichard, P. (1963) J. Biol. Chem. 238, 3407-3413. Tye, B. K., and Lehman, I.R. (1977) J. Mol. Biol. 117, 293-306. Brynolf, K., Elinsson, R., and Reichard, P. (1978) Cell 1;, 573-580. Lindahl, T., Ljungquist, S., Siegert, W., Nyberg, B., and Specens, B. (1977) J. Biol. Chem. 252, 3286-3294. DaRoza, R. Friedberg, E.C., Duncan, B.K., and Warner, H.R.,(1977) Biochemistry 14, 4934-4939. Thomas, K.R., Mahalapaz-Ramos, P., Lundquist, R., and Olivera, B.M. (1978) Cold Spring Harbor Symp. Quant. Biol. (in press). Karran, P., and Lindahl, T. (1978) J. Biol. Chem 253, 5877-5879. Riazuddin, S., and Lindahl, T. (1978) Biochemistry 11, 2110-2118. Deutsch, W.A. and Linn, S. (1979) Proc. Natl Acad. Sci. U.S.A. Z—6_.' 141-1440