Anim. Blood Grps biochem. Genet. 8 (1977): 149-156

Haemolytic anaemia in Basenji dogs. 2. Partial deficiency of erythrocyte pyruvate kinase (PIC; EC 2.7.1.40) in heterozygous carriers'

E. Andresen

Department of Animal Genetics, Royal Veterinary and Agricultural University, Copenhagen, Denmark

Received: 24 May 1977

Summary

Congenital nonspherocytic haemolytic anaemia (HA) in dogs of the Basenji breed is inherited as a simple, autosomal recessive trait. Previous results of pyruvate kinase (PK) assays suggest a causal relationship between the anaemia and PK defi- ciency in erythrocytes. In the present investigation assays of this enzyme have been performed on haemolysates from 45 Basenji dogs, comprising 3 anaemic and 42 non-anaemic individuals of which 13 were known heterozygotes. The PK activity in haemolysates from the 42 non-anaemic dogs exhibited a bimodal distribution corresponding to the genotypic classes: heterozygotes and normal homozygotes. The results indicate that heterozygotes have a partial, detectable enzyme deficiency, not reflected in clinical disease, and thus give evidence of a gene dosage effect in agreement with observations in man. The proposed genotypes PK PK, PK pk and pk pk refer to normal homozygotes, heterozygotes, and anaemic individuals, respec- tively. The findings strengthen the genetic hypothesis of recessiveness of the anae- mia by direct detection of heterozygosity of parents of affected individuals. More- over, the results are of value in comparative studies and they have practical appli- cation in connection with eradication programmes.

Introduction

Familial occurrence of haemolytic anaemia (HA) in dogs of the Basenji breed was first reported from the USA by Tasker et al. (1969) and Erwin (1969). The latter author called attention to a similarity between the symptoms observed in the Basenji dogs and the clinical and haematological manifestations of pyruvate kinase (PK) deficiency in man as first described by Valentine et al. (1961). Searcy et al. (1971)

This work was supported by the Carlsberg Foundation.

149

E. ANDRESEN

obtained further evidence of erythrocyte pyruvate kinase deficiency being responsi- ble for the anaemia, but Teng (1974) and Standerfer, Templeton & Black (1974) observed increased PK activity in haemolysates from anaemic Basenji dogs as compared with non-anaemic individuals. However, Teng (1974) and Brown & Teng (1975) detected decreased PK activity in erythrocytes from a number of first degree relatives of affected dogs. Such individuals were regarded as heterozygous carriers in agreement with observations in man. Although this assumption may be correct a different approach is required. That is, firstly, to assess the mode of inheritance of the anaemia based on family studies, and secondly, to evaluate the results of PK assays on individuals of known genotype. Results of the first part of this approach have shown that the anaemia is inherited as a simple, autosomal recessive trait (Andresen, 1977). Results of the pyruvate kinase assays are presented in this paper and the purpose is to relate these results to the results of the genetic investigations.

Material and methods

Animals The animals were registered by the Danish Kennel Club (DKK). They were from one to 13 years old when subjected to PK assay and they belonged to generations VIII to XV in the genealogical diagram shown in Fig. 1, but a few individuals be- longed to branches not shown in this diagram. The total material comprised 3 anaemic dogs (1 male and 2 females) and 42 non-anaemic dogs (16 males and 26 females). Distinction between anaemic and non-anaemic dogs was based on results of standard haematological tests. The non-anaemic dogs belonged to four classes A, B, C, and D, according to genetic criteria described under results and discussion (cf. Fig. 2).

Preparation of haemolysates Blood was collected into tubes with heparin. About lyi ml ACD solution was added per 5 ml of blood. The ACD solution contained 8 g citric acid, 22 g sodium citrate, 24.5 g glucose and distilled water ad 1000 ml. The samples were placed in ice until storage in refrigerator at 4 "C. One half volume of saline was added prior to centrifugation at 2100 g for 10 min. The supernatant and the buffy-coat was removed and to one volume of the packed cells was added about 1% vol. of saline. Following mixing c. 0.4 ml of the cell suspension was centrifuged at 3000 rev/min in a spiral centrifuge using polyethylene capillary tubing as test tubes (length 500 mm, diam. 1.5/1.0 mm). The centrifuge was made by Svend SchrZder, Copenhagen, according to specifications. Danielsson & Lundmark (1957) and Hanel (1969) have described the principles and the effect of this centrifugation procedure. After the spiral-centrifugation the supernatant white cells were removed by cutting the tube at the appropriate place. The packed red cells were transferred into a test tube, and after counting of the remaining white cells about 1 ml of H,O

150 Anim. Blood Grps biochem. Genet. 8 (1977)

HAEMOLYTIC ANAEMIA IN BASENJI DOGS. 2

IX

X

X I

X I 1

Xlll

X I V

x v

I

I I

111

IV

V

VI

Fig. 1. Genealogical diagram for Basenji dogs affected by haemolytic anaemia (HA). The proband litter in generation XV includes two females, H and I, with HA. A total of 22 HA dogs and 33 heterozygous carriers are indicated by respective symbols. The heterozygosity or" these individuals is based exclusively on genetic criteria (Andresen, 1977).

was added and the mixture was frozen at -20 "C until the next day. Samples of packed cells containing more than 1000 white cells per mm3 were rejected. After melting and centrifugation the haemolysate usually contained 2%-3 g Hb/lOO ml. Exact measurement was made using the Merckotest reagent kit from Merck, Ger- many.

Anim. Blood Grps biochem. Genet. 8 (1977) 151

E. ANDRESEN

Pyruvate kinase assay One unit of PK activity was defined as one micromole phosphoenolpyruvate con- verted (to pyruvate) per minute per gram haemoglobin at 37 "C. Beutler (1975) has described the principles and the essential steps of the procedure of the PK assay applied. Here it may suffice to mention that a test kit from Boehringer Mannheim, West Germany was used, and the change in optical density was recorded during a period of c. 20 min.

The units of activity are indicated with one decimal in Fig. 2 and all results are based on one assay per individual. Duplicate measurements were performed on blood samples from 10 individuals to estimate,the error of measurement as ex- pressed by the following estimates: mean difference = 0.01; standard deviation of the difference between the results of the paired measurements sD = 0.24, and

P K p k PKPK --

i e t e r o z y g o t e s : h o m o z y g o t e s

I.... I. .. D ( s e e legendl I . Activity: 0 0.5 1 .o 1.5 E

Fig. 2. Pyruvate kinase (PK) activity in erythrocytes from 42 non-anaemic Basenji dogs be- longing to four classes based on genetic criteria described in the text. The expected genotypic ratio of the 12 individuals in class D could not be ascertained. Solid squares and circles repre- sent activity in males and females, respectively. The histogram and associated curves of the PK activities indicate a bimodal distribution, deduced to represent heterozygous carrien (PK pk) and normal homozygotes (PK PK). The shaded area represents known heterozygotes, all be- longing to class A. The smooth curves were drawn after computing a number of function values as indicated in the figure.

152 Anim. Blood Grps biochem. Genet. 8 (1977)

HAEMOLYTIC ANAEMIA IN BASENJX DOGS. 2

standard deviation of single determinations s = sD/V2 = 0.17 (Snedecor & Cochran, 1967). The ten individuals exhibited a PK range of 0.3-1.6 units with an average of 0.8.

Results and discussion

Initial investigations confirmed the importance of an effective separation of red and white blood cells since the pyruvate kinase (PK) content of white cells is several hundred folds higher than that of red cells. Moreover, only the erythrocytic PK activity was expected to be associated with the haemolytic anaemia (HA), as it is normal in white cells in cases ol depressed activity in red cells. However, haemato- logical investigations showed that the three HA dogs among the total material of 45 dogs had only a few normal erythrocytes, which could not be separated from various kinds of white cells. Applying identical separation procedure as for non- anaemic individuals the three HA dogs exhibited PK activities of 13, 26, and 26 units, i.e. 10-50 times larger than otherwise found. Thus, the assumption that the HA condition is due to erythrocyte PK deficiency could not be examined directly.

The PK activity in the 42 non-anaemic dogs had a range of 0.3 - 1.8 units. These dogs were divided into four classes. A, B, C, and D (see Fig. 2) in order to relate the PK activity to hypothetical genotypes, i.e. PK PK for normal homozygotes and PK p k for heterozygotes, pk p k being the genotype for anaemic individuals. Class A comprises 13 non-anaemic parents of one or more offspring affected by haemolytic anaemia. Such parents must be heterozygous with regard to the haemolytic anaemia according to previous investigations already referred to (Andresen, 1977). Figure 2 indicates that these individuals exhibited a PK activity within the range of 0.3 to 1.1 units.

Class B comprises 8 non-anaemic litter mates of individuals with haemolytic anaemia. Thus, the expected ratio of heterozygous : normal homozygous individuals within this class is 2:l. The results show that 6 individuals exhibited PK activity within the range observed for the heterozygous individuals in class A, whereas 2 showed higher values, i.e. 1.2 and 1.3 units, respectively. It should be noticed, how- ever, that the tentative demarcation PK value of E = 0.99 (to be considered later) leads to a number of four presumed homozygotes in this class.

The C class comprises 9 non-anaemic individuals which are offspring of non- anaemic parents of which one is known to be heterozygous according to genetic criteria, whereas the other parent is either normal homozygous or heterozygous. Hence, the expected ratio of heterozygous : normal homozygous offspring is 1: (<l). The actual results show that 4 individuals exhibited PK activity within the range observed for the heterozygous individuals in class A, whereas 5 had higher values, i.e. 1.4 - 1.8 units.

Each individual, 12 in number, for which the probability of heterozygosity could not be ascertained was placed in class D.

Anim. Blood Grps biochem. Genet. 8 (1977) 153

E. ANDRESEN

The 13 known heterozygotes in class A have a mean activity of x = 0.60 and a standard deviation of s = 0.23. This indicates that less than 5 % of normally dis- tributed activity values from heterozygous individuals are expected to exceed the value of x -t 1.645 s = 0.99, the figure 1.645 being the appropriate standard normal deviate (Snedecor & Cochran, 1967). Thus, this value of 0.99 may be regarded as a tentative demarcation between presumed heterozygous and normal homozygous individuals. Using this demarcation value the observed ratios of pre- sumed heterozygotes : homozygotes in classes B and C are in agreement with the expectations indicated in Fig. 2. Hence, for statistical purposes, the two genotypes in the total material comprise 30 and 12 individuals, respectively, although on2 member of the latter group is known to be heterozygous, i.e. the individual in class A exhibiting an activity value of 1.1. The 30 presumed heterozygous individuals have an average activity of 0.63 and a standard deviation of 0.18, and the 12 pre- sumed normal homozygous individuals have an average activity of 1.32 and a standard deviation of 0.24.

The activities are plotted on a histogram which clearly shows a bimodal distri- bution. The height of each column corresponds to the respective number of indivi- duals within each interval. The 13 known heterozygotes correspond to the shaded area. The continuous curves are fitted using the relevant numbers, means, and standard deviations just considered. In the figure are indicated two black areas corresponding to 5 % tail areas within the overlapping distributions. The vertical demarcation for the two areas correspond to activity values of 0.92 and 0.93 which supports the choice of the value of 0.99 to separate presumed heterozygotes from normal homozygotes.

The mean values and associated 95 % confidence intervals for the two distri- butions are 0.63 2 0.06 and 1.32 k 0.15, respectively. Thus, the results give evidence of an additive dosage effect of the normal gene for PK activity and they support the assumption of a causal relationship between PK deficiency and haemo- lytic anaemia, heterozygous carriers being partially deficient but clinically normal. The term deficiency in this context merely refers to decreased PK activity in hetero- zygotes under the test conditions since the procedure does not lead to a distinction between various causes of the decreased activity (see e.g. Brandt & Hanel, 1971; Standerfer et al., 1974; Standerfer et al., 1975; Adachi et al., 1977).

The conclusion of the present investigation is based on the assumption that only one locus with two alleles is involved. Meanwhile it should be noticed that the known heterozygous carriers in class A exhibit a PK activity range of 0.3 - 1.1 units with an average activity of 0.60, whereas the range of presumed heterozygous in- dividuals in the classes B and C is 0.6 - 0.9 units with an average of 0.76, all values being larger than the average of class A. The tail of low values in class A may sug- gest a correlation between PK values and age of the individuals, but the data do not support this assumption. Furthermore, the possibility that the tailing is as- sociated with multiple alleles is not supported by genetic investigations of the same

154 Anirn. Blood Grps biochern. Genet. 8 (1977)

HAEMOLYTIC ANAEMIA IN BASENJI DOGS. 2

material (Andresen, 1977, cf. Fig. 1). Finally, among the 12 non-anaemic indivi- duals in class D, 10 had PK values ranging from 0.3 to 0.9 units with an average of 0.60. Hence, the absence of values less than 0.6 in classes B and C is probably fortuitous.

The degree of overlapping of the two normal distribution curves in Fig. 2 is an indication of the discriminatory power of PK assays in detecting heterozygous car- riers. Moreover, the importance of repeated PK assays of an individual is clear from the magnitude of the error of measurement indicated under material and methods. The PK activity corresponding to the interception of the curves may be regarded as a guideline value in eradication programmes. In the present situation this value is between 0.95 and 1.0. Doubtless, a very effective selection will be achieved if individuals having erythrocytic PK activity less than this value are excluded from breeding. It should be noticed that this value may vary between laboratories, e.g. depending on the chosen method, the temperature of the assay and whether the PK activity is measured relative to the amount of haemoglobin or number of erythrocytes per assay.

Addendum. The data presented are clearly in agreement with the hypothesis of an additive gene effect, according to which the mean activity of heterozygous indivi- duals should be half of the one found in normal homozygotes. This is confirmed by preliminary computations using a computer programme for fitting distribution mixtures, being developed by J. Hilden.

Acknowledgments

Thanks are due to Dr H. K. Hanel, Department of Clinical Chemistry, Bispebjerg Hospital, Copenhagen, and to Dr J. Hilden, Institute of Human Genetics, University of Copenhagen, for valuable advice and discussion of the subject (see also ‘Adden- dum’). Mrs Inger Hvidtfeldt Christensen is acknowledged for excellent assistance. The owners of various dogs and Danish Kennel Club kindly provided information about each pedigree.

References

Adachi, K., P. K. Ghory, T. Asakura & E. Schwartz, 1977. A monomeric form of pyruvate

Andresen, E., 1977. Haemolytic anaemia in Basenji dogs: I. Genetic investigations. Hereditas

Beutler, E., 1975. Red cell metabolism, 2nd ed. Grune & Stratton, London. Brandt, N. J. & H. K. Hanel, 1971. Atypical pyruvate kinase in a patient with haemolytic

Brown, R. V. & Yao-sheng Teng, 1975. Studies of inherited pyruvate kinase deficiency in the

Danielsson, H. & F. Lundmark, 1957. En ny typ af centrifug (spiralcentrifug). (The spiral

155

kinase in human pyruvate kinase deficiency. Proc. Nut1 Acad. Sci. 74 (2): 501-504.

85’: 211-214.

anaemia. Scnnd. J. Hnemntol. 8 : 126-133.

Basenji. 1. Am. Anim. Hosp. Ass. 11: 362-365.

centrifuge. A new type of centrifuge). Nord. Med. 31: 175.

Anim. Blood Grps biochem. Genet. 8 (1977)

E. ANDRESEN

Erwing, G. O., 1969. Familial nonspherocytic hemolytic anemia of Basenji dogs. 1. A m . vet.

Hanel, H. K., 1969. Pyruvate kinase in human erythrocytes. Scand. 1. Haernatof. 6: 173-174. Searcy, G. P., D. R. Miller & J . B. Tasker, 1971. Congenital hemolytic anemia in the Basenji

dog due to erythrocyte pyruvate kinase deficiency. Can. 1. cornp. Med. 35: 67-70. Snedecor, G. W. & W. G. Cochran, 1967. Statistical methods, 6th ed. The Iowa State University

Press, Iowa, USA. Standerfer, R. J., J. W. Templeton & J. A. Black, 1974. Anomalous pyruvate kinase deficiency

in the Basenji dog. 1. Am. vet. Med. Ass. 35: 1541-1543. Standerfer, R. J., M. B. Rittenberg, C. J. Chern, J. W. Templeton & J. A. Black, 1975. Canine

erythrocyte pyruvate kinase. 11. Properties of the abnormal enzyme associated with hemolytic anemia in the Basenji dog. Biochem. Genet. 13: 341-351.

Tasker, J . B., G. A. Severin, S. Young & E. L. Gillette, 1969. 1. A m . Vet. Med . Ass. 153:

Teng, Yao-sheng, 1974. Congenital nonspherocytic hemolytic anemia due to pyruvate kinase deficiency in Basenji dogs. Ph. D. Dissertation, Univ. of Missouri, Columbia, Mo.

Valentine, W. N., K. R. Tanaka & S. Miwa, 1961. A specific erythrocyte glycolytic enzyme defect (pyruvate kinase) in three subjects with congenital non-spherocytic hemolytic anemia. Trcins. Ass. A m . Physicians 74: 100-110.

Med. Ass. 154: 503-507.

158-165.

156 Anim. Blood Grps biochem. Genet. 8 (1977)