- Email: [email protected]
Some aspects of erythrocyte metabolism in insulin-treated diabetic dogs
Research in Veterinary Science 2002, 72, 23±27 doi:10.1053/rvsc.2001.0515, available online at http://www.idealibrary.com on
Some aspects of erythrocyte metabolism in insulin-treated diabetic dogs S. COMAZZI, S. PALTRINIERI, V. SPAGNOLO, P. SARTORELLI Dipartimento di Patologia Animale, Igiene e SanitaÁ Pubblica Veterinaria, Sezione di Patologia Generale Veterinaria e Parassitologia, UniversitaÁ degli Studi, Milan, Italy SUMMARY Insulin-dependent diabetes mellitus (IDDM) is a common metabolic disease often complicated by a number of pathological conditions among which are haematological changes and alterations in blood cell function. Human and feline diabetes mellitus patients have been reported to be associated with oxidative stress that can lead to membrane alterations and to reduced erythrocyte life-span. Erythrocyte function in dogs affected by IDDM has been investigated during insulin therapy, paying attention to antioxidant status, membrane resistance, enzyme activities and 2,3-diphosphoglycerate (2,3DPG) concentration. Thirteen diabetic and 36 healthy dogs were bled and haematology and blood chemistry assays were performed to evaluate the degree of compensation. Osmotic fragility, the activities of the enzymes glucose-6-phosphate dehydrogenase (G6PD) and pyruvate-kinase (PK) and the concentrations of reduced glutathione (GSH) and 2,3DPG were evaluated in the erythrocytes. Diabetic dogs did not differ from controls in terms of haematological parameters, except for higher numbers of platelets. Higher values of fructosamine, glucose, protein, plasma potassium and calculated osmolality were detected in the plasma from diabetic dogs. No differences were detected in osmotic fragility, GSH concentration and PK activity between the two groups but 2,3DPG concentration and G6PD activity were statistically significantly higher in the diabetic group. The results indicate minimal alterations in erythrocyte functions occur in insulin-treated diabetic dogs. This contrasts with what has been reported for IDDM humans and cats. # 2002 Harcourt Publishers Ltd
INSULIN-DEPENDENT diabetes mellitus (IDDM) is one of the most common metabolic disorders occurring in dogs with a reported incidence as high as 15 per 1000 (Kaneko 1997). Haematological complications commonly develop in human diabetes mellitus. Erythrocyte morphology, functions and metabolism can be influenced by three general mechanisms: glycosylation of membrane proteins and enzymes, oxidative damage mainly due to glucose auto-oxidation, and changes related to ketoacidosis (Christopher 1995). These mechanisms cause haemoglobin glycosylation and altered oxygen affinity, decreased membrane deformability and alterations of metabolic patterns. They all contribute to shorten erythrocyte lifespan or to decreased oxygen availability to the tissue. Erythrocyte metabolism is mostly restricted to two main pathways. The Embden-Meyerhof pathway (EMP) is mainly involved in ATP and 2,3DPG generation in the erythrocytes, while the pentose phosphate pathway (PPP) generates NADPH, as a source of reducing equivalents to protect erythrocytes from oxidation. The activity of the enzymes pyruvate kinase (PK) and glucose-6-phosphate dehydrogenase (G6PD) are often tested as key-enzyme of EMP and PPP, respectively. Corresponding author: Dr Comazzi, Dipartimento di Patologia Animale, Igiene e SanitaÁ Pubblica Veterinaria, Sezione di Patologia Generale Veterinaria e Parassitologia, UniversitaÁ degli Studi, Via Celoria 10, 20133 Milan, Italy. Tel.: 0039-02-26680443; Fax: 0039-02-2364470; E-mail: [email protected]
0034-5288/02/010023 05 $35.00/0
In humans glycosylation affects the function of haemoglobin binding to 2,3-diphosphoglycerate (2,3DPG) site, thus influencing the regulation of the affinity with oxygen (Jones and Peterson 1981). However, in human medicine, findings concerning 2,3DPG concentration in the erythrocytes during diabetes mellitus are contradictory (Ditzel and Standl 1975, Kanter et al 1975). Decreases in G6PD activity are reported in human patients affected by complicated diabetes mellitus (Costagliola et al 1988, Costagliola 1990), contributing to increase erythrocyte susceptibility to oxidative stress. However, other authors have reported an increase in G6PD activity in uncontrolled diabetics (Goebel et al 1975). During diabetes mellitus many membrane changes due to oxidative stress are responsible for a decrease of osmotic resistance and of shortened erythrocyte lifespan (Gandhi and Chowdhury 1979, Jain 1989). Reduced glutathione (GSH) is one of the most important protecting agents by means of oxidation to a disulfide (GSSG). GSSG is reduced to GSH by glutathione reductase, using NADPH as a substrate to restore GSH concentration. Decreases in GSH concentration in diabetes mellitus have been reported both in human patients (Gandhi and Chowdhury 1979, EI Fakhri et al 1987, Konukoglu et al 1999) and in ketoacidotic cats (Christopher 1989, Christopher 1995). This has been associated with Heinz bodies formation (Christopher 1989). There are no reports regarding erythrocyte functions in canine diabetes mellitus. The haematological # 2002 Harcourt Publishers Ltd
24
S. Comazzi, S. Paltrinieri, V. Spagnolo, P. Sartorelli
complications during insulin therapy are very important because they can strongly influence both treatment and outcome in canine IDDM. In this study osmotic fragility, the activities of PK and G6PD and the concentrations of 2,3DPG and GSH have been investigated in order to better clarify the alterations of some erythrocyte functions in dogs affected by diabetes mellitus during insulin therapy. MATERIALS AND METHODS Blood samples were collected from 13 dogs affected by diabetes mellitus, at least 6 months from beginning insulin therapy, and from 36 healthy dogs that served as a control group. Insulin therapy had been chosen by the clinicians and administered once daily at the morning. Both well and poorly compensated dogs were chosen, but none of them showed clinical signs consistent with haematological or vascular complications. Five millilitres of blood were collected into lithium heparin tubes and 05 ml of blood was added to citratedextrose solution (ACD) anticoagulated tubes for GSH determination. The samples were immediately refrigerated at 4 C and all the tests on the erythrocytes were assayed within 2 hours. A complete blood cell count was performed by means of an automated analyser (H8 SEAC, Calenzano FI, Italy), and a differential leukocyte count was performed on a May-GruÈnwald Giemsa stained smear. Reticulocytes and Heinz bodies were evaluated on a brillant-cresyl blue stained smear (Pasquinelli 1984). Plasma was immediately separated and stored at ÿ30 C, and the samples were analysed for chemical assays within 2 weeks. Glucose (GOD-POD method, Hospitex Diagnostic, Firenze, Italy), total protein (biuret reaction, Hospitex Diagnostic, Firenze, Italy), urea nitrogen (enzymatic method, Hospitex Diagnostic, Firenze, Italy), albumin (bromocresol green colorimetric method, Hospitex Diagnostic, Firenze, Italy), inorganic phosphate (ammonium molybdate colorimetric assay, Hospitex Diagnostic, Firenze, Italy), fructosamine (nitroblue tetrazolium reduction, SEAC, Firenze) and b-hydroxybutyrate (Sigma Diagnostic, St Louis, US) were assayed on an automated analyser (Hospitex Diagnostic, Firenze, Italy) using commercial kits. Globulin was calculated subtracting albumin from total protein concentrations. Sodium and potassium were evaluated by a flame photometer (FP20, SEAC, Calenzano FI, Italy). Fibrinogen was estimated by means of a refractometer (Jain 1986). Osmolality was calculated as described by Carlson (1997). Osmotic fragility A 03 ml aliquot of blood was used to test osmotic fragility, according to the method of Dacie and Lewis (Miale 1982). Briefly, 25 ml of blood were mixed with 25 ml of the following concentrations of NaCl: 09 per cent; 075 per cent; 065 per cent; 055 per cent; 05 per cent; 045 per cent; 04 per cent; 035 per cent; 03 per cent; 02 per cent; 01 per cent. After an incubation
of 30 minutes at room temperature and centrifugation at 350 g for 10 minutes, the supernatants were read at a wavelength of 540 nm against water. 09 per cent NaCl was used as a negative control and the 01 per cent NaCl as a positive control from which complete haemolysis was obtained. From the values obtained, the fragiligram, composed of a cumulative (dose-response sigmoid curve) and a derivative curve (progressive haemolytic increment) of haemolysis, was drawn, and the concentrations of NaCl corresponding to the minimum (less than 5 per cent haemolysis), medium (50 per cent haemolysis) and maximum (more than 90 per cent haemolysis) osmotic fragility were calculated. 2,3DPG concentration An aliquot (05 ml) of blood was deproteinized in 15 ml of cold 8 per cent trichloroacetic acid. After centrifugation, the supernatant was stored at ÿ20 C and the 2,3DPG concentration was analysed within a month using a commercial kit (Sigma Diagnostic, St Louis, MO, USA), on a spectrophotometer at 340 nm. Haemoglobin concentration was tested by means of Drabkin's method (SEAC H10, Firenze, Italy) and the concentration of 2,3DPG was expressed in mmol gÿ1 Hb. Enzyme activities An aliquot (14 ml) of blood was centrifuged (15 minutes at 350 g) and plasma and buffy coat were carefully removed by aspiration. PK and G6PD activities were measured at 340 nm and 25 C in the haemolysate using commercially available kits (Boehringer Mannheim, Mannheim, Germany). Enzyme activities were expressed in mU gÿ1 Hb. Reduced glutathione (GSH) GSH concentration was evaluated spectrophotometrically at 412 nm, after deproteinisation with metaphosphoric acid (167 per cent) and EDTA (02 per cent) solution, using 5.5 0 -dithiobis-(2-nitrobenzoic acid) (Pasquinelli 1984). GSH concentration was expressed as mg/g Hb.
Statistical analysis Statistical analysis was performed by mean of a statistic software (Statistica, Stat Soft Inc, Tulsa, OK, USA). The t test for normally distributed data and the Mann-Whitney U test for nonparametric distributions were used. Correlations between haematology and chemical results and erythrocyte function were performed by the Spearman test. Results were considered significant when P < 005.
RESULTS The results of haematology and blood chemistry, which were significantly different between diabetic dogs
25
Erythrocytes metabolism in diabetic dogs TABLE 1: Haematology and blood chemistry results (mean and standard deviation), including number of cases and statistically significant differences between diabetic and control groups
Platelets n mlitre ÿ1 Total Protein g litre ÿ1 Globulin g litre ÿ1 Estimated fibrinogen mg dlitre ÿ1 Calculated osmolality mOsm litre ÿ1 Glucose mmol litre ÿ1 Fructosamine mmol litre ÿ1 Plasma potassium mmol litre ÿ1
Diabetic Mean (S.D.)
n
Controls Mean (S.D.)
n
P5
314917 (156877) 827 (117) 439 129 031 (021)
12
177077 (76586) 672 (78) 329 54 017 (013)
13
001
21
0001
21
001
17
005
12 12 10
341 (33)
11
316 (21)
13
005
1976 (1282) 434 (150) 481 (039)
12
538 (150) 291 (73) 438 (044)
13
0001
13
001
13
005
13 12
TABLE 2: Enzyme activities, 2,3DPG and GSH concentrations and osmotic fragility results (mean and standard deviation), number of cases and statistically significant differences between diabetic and control group; n.s. not significant
GSH /Hb
mg g ÿ1 Hb
PK
mU gÿ1 Hb G6PD mU gÿ1 Hb 23 DPG mmol g ÿ1 Hb min frag % NaCl med frag % NaCl max frag % NaCl
Diabetic Mean (S.D.)
n
Controls Mean (S.D.)
n
P5
234 (091) 161 (062) 491 (143) 1791 (465) 0486 (0060) 0414 (0040) 0322 (0061)
13
275 (098) 283 (217) 340 (120) 1437 (331) 0527 (0095) 0445 (0058) 0357 (0071)
13
n.s.
36
n.s.
36
0001
36
001
36
n.s.
36
n.s.
36
n.s.
13 13 12 13 13 13
DISCUSSION and the control group are shown in Table 1. No alterations in erythrocyte number, haemoglobin concentration, PCV, leukocyte counts and erythrocytic indexes were detected. Only higher platelet number was found (P < 001) in the diabetic dogs. No differences in the number of reticulocytes and nucleated RBCs were found, and no Heinz bodies were detectable in the blood from both groups. A slight lipaemia was noticed in some of serum samples even if no strong lipamic samples were found. Blood chemistry showed an increase of glucose and fructosamine in the diabetic group. Total protein and globulins were higher in the diabetic dogs as well as estimated fibrinogen without increases in albumin concentration. Calculated osmolality and plasma potassium were higher in the diabetic group. No alterations of b-hydroxybutyrate concentration were detected in the dogs examined. The results of osmotic fragility, enzymes activities and the concentrations of 2,3DPG and GSH are shown in Table 2. No statistical differences were detected in the activity of PK of the two groups but the mean value and the standard deviation were lower in the diabetic dogs while the distribution of the results in the control group was much wider. In contrast higher activities (P < 0001) of G6PD were detected in the diabetic group. No differences were found in GSH concentration but the mean value was lower in the diabetic group. In contrast diabetic dogs exhibited higher concentrations of 2,3DPG (P < 001) than control group. No differences were detectable in the osmotic fragility between the two groups. A positive correlation between plasma glucose concentrations and the activities of G6PD (P < 005, r 0664) and a negative correlation between the activities of G6PD and the concentrations of GSH in the erythrocytes (P < 005, r ÿ0626) were found.
The pathogenesis of canine diabetes mellitus strongly differs from feline and human maturity-onset forms and seems to be more similar to juvenile-onset diabetes in humans. The haematological complications of diabetes mellitus have been widely investigated in non-insulin dependent diabetes mellitus (NIDDM) of human beings (Jones and Peterson 1981) and partly in feline diabetes (Christopher 1989, Christopher 1995). However, no data are available about erythrocyte metabolism and functions in canine diabetes mellitus. Our results suggest few alterations of the haematological pattern occurs in insulin-treated diabetic dogs. Only platelets counts were higher in the diabetic dogs than in controls, but were still in the normal reference range. Anaemia has been reported more frequently in non-survivor diabetic dogs in comparison with survivor ones (Ling et al 1977), but this can sometimes be partly masked by dehydration associated with hyperglycaemia and hyperketonaemia (Christopher 1995). Anaemia was not found in any of our diabetic dogs and even if we cannot exclude dehydration as suggested by hyperosmolality in the diabetic group, was probably related to compensation induced by insulin therapy or to lesser numbers of damaged erythrocytes. Blood chemistry showed higher values of plasma glucose and fructosamine concentration in diabetic dogs, as was expected. Fructosamine is a very useful parameter to monitor the response to insulin treatment in diabetic dogs, because it reflects the average plasma glucose concentration over the preceding two weeks, and it is not affected by acute changes in the glucose concentration (Jensen 1995). In our dogs fructosamine concentrations fell into the range reported by Jensen (1995) for diabetic dogs (326±12429 mmol litreÿ1) and only one sample had a high concentration of fructosamine (788 mmol litreÿ1) while none of the dogs had ketosis. The glucose and fructosamine results showed a wide variation within the
26
S. Comazzi, S. Paltrinieri, V. Spagnolo, P. Sartorelli
population mainly because of one extreme outlying value. However, most values were higher than reference values indicating there was only partial regulation of diabetes in the group of dogs we examined. This presence of a high number of poorly controlled diabetic dogs together with the results of the correlation tests, suggest that the observed alterations of RBC metabolism are probably related to the severity of diabetes. Some evaluation in vitro to identify the direct effects of hyperglycaemia on enzyme activities and 2,3DPG concentration would be of interest. Plasma proteins, as well as fibrinogen, were higher in the diabetic group than in controls. This is due to an increase of globulins as no differences in albumin concentrations were detectable. Hyperfibrinogenaemia has been reported to be often related to diabetes mellitus both in human (Le Devehat 2000) and in veterinary species (Christopher 1995), however the magnitude of hyperfibrinogenemia in our samples is insufficient to account for the degree of hyperproteinemia. Moreover, even if no gross lipaemic samples were found in the present study it is possible some slightly lipaemic samples could have falsely raised the concentration of total protein, as lipaemia has been reported to interfere with some haematochemical and refractometer assays (Alleman 1990). A moderate increase in plasma potassium was detected in the blood collected from diabetic dogs with 5 out of 12 samples being higher than 50 mmol litreÿ1. Cotton et al (1971) reported hyperkalaemia in 29 per cent and hypokalaemia in 14 per cent of a sample of 28 diabetic dogs. In their sample, however, both well and poorly controlled diabetics were present and 11 of 28 dogs did not survive on insulin therapy. Hypokalaemia is often related to poorly managed insulin therapy and it is one of the most serious problems concerning insulin therapy and ketoacidosis (Nichols and Krenshaw 1995). When only considering the surviving dogs, the percentage of dogs with hyperkalaemia in Cotton's study increased to 39 per cent, which is similar to what we observed. The absence of hypokalaemia in the dogs we tested confirm the partial regulation of diabetes. In human diabetics an increase in osmotic fragility (Gandhi and Chowdhury 1979, Jain 1989) and a decrease in GSH concentration (Gandhi and Chowdhury 1979, Konukoglu et al 1999, Zaltzberg et al 1999) have been reported, and is often associated with diabetic complications (Thornalley et al 1996). Jain (1989) reported the increase of lipid peroxidation as one of the main causes of increased osmotic fragility of human erythrocyte in vitro. The dogs we observed did not exhibit any alteration of osmotic fragility or statistically significant decreases in GSH concentration. This could be related to a less damage to membrane structures in insulin-treated diabetic dogs, as suggested also by the absence of anaemic dogs in our samples. G6PD activity increased in diabetic dogs when measured in vitro. In vivo the activity of G6PD is regulated by many different conditions: changes in pH, the concentration of substrates (NADP and glucose-6 phosphate), inhibitors (nucleotides, metal ions, hormones, intermediates of glycolytic and polyol pathways) and activators (Ca, Mn, Mg). The activities of
G6PD in vitro were tested under optimal conditions, at stable pH and with excess NADP , but without other inhibitors or activators being added. Mature erythrocytes cannot synthesise new enzyme. Even though in vitro G6PD activity far exceeds the activity that occurs within the erythrocytes in vivo, we cannot exclude the possibility that in vitro increases of G6PD activities in diabetic dogs could reflect an in vivo upregulation of the enzyme in response to diabetic complications. The relation between this increase in activity and the severity of diabetes is suggested by the positive correlation between G6PD activities and glucose concentration and by the negative correlation between the G6PD activities and the concentration of GSH. Our results, obtained in insulin- treated patients, are partly in agreement with those obtained in humans, in which an increased activity of G6PD has been found in diabetic patients (Goebel et al 1975). On the other hand, other authors (Costagliola 1990) reported a decrease of G6PD activity in diabetic human patients, which might affect the integrity of the glutathione system. 2,3DPG concentration increased in the diabetic group. In previous studies, 2,3DPG concentration has been reported to increase in hypoxic circumstances due to cardiomyopathy, anaemias and polycythaemia vera (Paltrinieri et al 1998, Paltrinieri et al 2000, Comazzi et al 2000). In human affected by diabetes mellitus oxygen affinity has been reported to be abnormal, concomitant with the glycosylation of the 2,3DPG binding site or with some vascular diseases (Jones and Peterson 1981), and the concentration of 2,3DPG is reported to be markedly decreased in diabetic ketoacidosis (Ditzel and Standl 1975) and in poorly controlled diabetes (Triolo et al 1981). This is due to a lower plasma concentration of inorganic phosphate. In our samples, no alteration of inorganic phosphate were detectable. In well controlled patients or in patients with vascular complications, hypoxia is reported in association with an increase in 2,3DPG concentration thus decreasing the affinity of haemoglobin with oxygen (Ditzel and Standl 1975, Kanter et al 1975, Samaja et al 1982). In contrast, some authors (Marschner et al 1994, Marschner and Rietbrock 1994) have reported that 2,3DPG increases may have only slight effects on oxygen release in human diabetic patients. Canine glycosylated haemoglobin (Hb Can A0) has been described by Enoki and others (1986) with abnormal oxygenation properties, which is slightly different from the human form. In particular, a marked depression in the 2,3DPG and CO2 effects was reported and this could suggest a poor compensation to tissue hypoxia. However, according to previous studies in different hypoxic conditions (Paltrinieri et al 1998, Paltrinieri et al 2000, Comazzi et al 2000), we cannot exclude that the increases of 2,3DPG could represent a poor attempt to compensate for tissue hypoxia due to vascular complications and altered oxygenation properties of glycosylated haemoglobin in diabetic dogs. In conclusion the results of this study indicate minimal evidence of oxidative damage in erythrocytes from dogs affected by insulin-treated diabetes mellitus, compared to feline and human diabetes. In particular
Erythrocytes metabolism in diabetic dogs
no alterations in osmotic fragility nor Heinz bodies formation were found while the concentration of natural antioxidant, glutathione, is similar to healthy dogs. The evaluation of enzymatic activities in vitro showed higher activities of G6PD but, even if they showed correlation with glucose and GSH concentrations, more studies are needed to determine if they are directly related to diabetic complications. A higher concentration of 2,3DPG suggests a response to hypoxic condition that could be related both to haematological (altered affinity for oxygen) and non-haematological (vascular or thrombotic) complications. ACKNOWLEDGEMENTS Supported by MURST (ex 60 per cent). The authors wish to thank Dr Monica Merlini for technical assistance. REFERENCES ALLEMAN, A. R. (1990) The effects of hemolysis and lipemia on serum biochemical constituents. Veterinary Medicine 85, 1272±1284 CARLSON, G. P. (1997) Fluid, electrolyte and acid-base balance. In: Kaneko JJ, Harvey JW, Bruss ML eds. Clinical biochemistry of domestic animals. 5th ed. San Diego: Academic Press; 485±516 CHRISTOPHER, M. M. (1989) Relation of endogenous Heinz bodies to disease and anemia in cats: 120 cases. Journal of the American Veterinary Medical Association 194, 1089±1095 CHRISTOPHER, M. M. (1995) Hematologic complications of diabetes mellitus. Veterinary Clinics of North America: Small Animal Practice 25, 625±637 COMAZZI, S., PALTRINIERI, S., AGNES, F., SACCHET, A. & MILANI, F. (2000) Some aspects of erythrocyte metabolism in a dog with polycythaemia vera. Veterinary Record 147, 331±334 COSTAGLIOLA, C. (1990) Oxidative state of glutathione in red blood cells and plasma of diabetic patients: in vivo and in vitro study. Clinical Physiology & Biochemistry 8, 204±210 COSTAGLIOLA, C., IULIANO, G., MENZIONE, M., NESTI, A., SIMONELLI, F. & RINALDI, E. (1988) Systemic human disease as oxidative risk factors in cataractogenesis. Ophtalmic Research 20, 308±316 COTTON, R. B., CORNELIUS, L. M. & THERAN, P. (1971) Diabetes mellitus in the dog: a clinicopathologic study. Journal of the American Veterinary Medical Association 159, 863±870 DITZEL, J. & STANDL, E. (1975) The problem of tissue oxygenation in diabetes mellitus. Acta Medica Scandinavica 578 suppl., 59±69 EL FAKHRI, M., SHERIFF, D. S., CHANDRASENA, L., GUPTA, J. D., RAI, S. T. & QURESHI, M. S. (1987) Erythrocyte glutathione concentration in diabetics with cataracts, with and without glucose-6-phosphate dehydrogenase deficiency. Clinical Chemistry 33, 1936±1937 ENOKI, Y., OHGA, Y., SAKATA, S., KOHZUKI, H. & SHIMIZU, S. (1986) Structural and functional characterization of a canine minor glycosylated hemoglobin, Hb Can A1c. Hemoglobin 10, 607±621 GANDHI, C. R. & CHOWDHURY, D. R. (1979) Effect of diabetes mellitus on sialic acid and glutathione content of human erythrocytes of different ages. Indian Journal of Experimental Biology 17, 585±587 GOEBEL, K. M., GOEBEL, F. D., NEITZERT, A., HAUSMANN, L. & SCHNEIDER, J., (1975) Adaptation of red cell enzymes and intermediate in metabolic disorders. Enzyme 19, 201±211
27
JAIN, N. C. (1986) Schalm's veterinary hematology. 4th ed. Philadelphia, Pa: Lea & Febiger, 58±59 JAIN S. K. (1989) Hyperglycemia can cause membrane lipid peroxidation and osmotic fragility in human red blood cells. The Journal of Biological Chemistry 264, 21340±21345 JENSEN, A. L. (1995) Glycated blood proteins in canine diabetes mellitus. Veterinary Record 137, 401±405 JONES, R. L. & PETERSON, C. M. (1981) Hematologic Alterations in diabetes mellitus. American Journal of Medicine 70, 339±352 KANEKO, J. J. (1997) Carbohydrate metabolism and its disease. In: Kaneko, J.J., Harvey J. W. & Bruss M. L. eds. Clinical biochemistry of domestic animals. 5th ed. San Diego: Academic Press; 45±81 KANTER, Y, BESSMAN, S. P. & BESSMAN, A. (1975) Red cell 2,3-diphosphoglycerate levels among diabetic patients with and without vascular complications. Diabetes 24, 724±729 KONUKOGLU, D., AKCAY, T., DINCER, Y. & HATEMI, H. (1999) The susceptibility of red blood cells to autoxidation in type 2 diabetic patient with angiopathy. Metabolism 48, 205±299 LE DEVEHAT, C., KHODABANDEHLOU, T. & VIMEUX, M. (2000) Diabetes mellitus and fibrinogen. Hemorrheological and microcircolatory consequences. Journal des Maladie Vasculaires 25, 53±57 LING, G. V., LOWENSTINE, L. J., PULLEY, T. & KANEKO, J. J. (1977) Diabetes mellitus in dogs: a review of initial evaluation, immediate and long term management and outcome. Journal of the American Veterinary Medical Association 170, 521±530 MARSCHNER, J. P. & RIETBROCK, N. (1994) Oxygen release kinetics in healthy subjects and diabetic patients. I: the role of 2,3-diphosphoglycerate. International Journal of Clinical Pharmacology & Therapeutics 32, 533±535 MARSCHNER, J. P., SEIDLITZ, T. & RIETBROCK, N. (1994) Effect of 2,3diphosphoglycerate on O2-dissociation kinetics of hemoglobin and glycosylated hemoglobin using the stopped flow technique and an improved in vitro method for hemoglobin dissociation. International Journal of Clinical Pharmacology & Therapeutics 32, 116±121 MIALE, J. B. (1982) Laboratory medicine: hematology. 6th ed. St Louis: The C. V. Mosby Company, 883±884 NICHOLS, R. & CRENSHAW, K. L. (1995) Complications and concurrent disease associated with diabetic ketoacidosis and other severe forms of diabetes mellitus. Veterinary Clinics North American: Small Animal Practice 25, 617±624 PALTRINIERI, S., COMAZZI, S. & AGNES, F. (2000) Haematological parameters and altered erythrocyte metabolism in anaemic dogs. Journal of Comparative Pathology 122, 25±34 PALTRINIERI, S., SARTORELLI, P., DE VECCHI, B. & AGNES, F. (1998) Metabolic findings in the erythrocytes of cardiopathic and anaemic dogs. Journal of Comparative Pathology 118, 123±133 PASQUINELLI, F. (1984) Diagnostica e tecniche di laboratorio. Firenze, Italy: Rosini editore, 875±876 & 914±915 SAMAJA, M., MELOTTI, D., CARENINI, A. & POZZA G. (1982) Glycosylated haemoglobins and the oxygen affinity of whole blood. Diabetologia 23, 399±402 THORNALLEY, P. J., MCLELLAN, A. C., LO, T. W., BENN, J. & SONKSEN, P. H. (1996) Negative association between erythrocyte reduced glutathione concentration and diabetic complications. Clinical Science 91, 575±582 TRIOLO, G., GIARDINA, E., CONTORNO, A., CIPOLLA, T., GIANNUOLI, G., CAIMI, G., CATANIA, A. & RINALDI, A. et al. (1981) Glycosylated hemoglobins and erythrocyte 2,3-diphosphoglycerate in diabetes mellitus. Archivio per le Scienze Mediche 138, 179±185 ZALTZBERG, H., KANTER, Y., AVIRAM, M. & LEVY, Y. (1999) Increased plasma oxidizability and decreased erythrocyte and plasma antioxidative capacity in patients with NIDDM. Israel Medical Association Journal 1, 228±231
Accepted October 10, 2001
Some aspects of erythrocyte metabolism in insulin-treated diabetic dogs S. COMAZZI, S. PALTRINIERI, V. SPAGNOLO, P. SARTORELLI Dipartimento di Patologia Animale, Igiene e SanitaÁ Pubblica Veterinaria, Sezione di Patologia Generale Veterinaria e Parassitologia, UniversitaÁ degli Studi, Milan, Italy SUMMARY Insulin-dependent diabetes mellitus (IDDM) is a common metabolic disease often complicated by a number of pathological conditions among which are haematological changes and alterations in blood cell function. Human and feline diabetes mellitus patients have been reported to be associated with oxidative stress that can lead to membrane alterations and to reduced erythrocyte life-span. Erythrocyte function in dogs affected by IDDM has been investigated during insulin therapy, paying attention to antioxidant status, membrane resistance, enzyme activities and 2,3-diphosphoglycerate (2,3DPG) concentration. Thirteen diabetic and 36 healthy dogs were bled and haematology and blood chemistry assays were performed to evaluate the degree of compensation. Osmotic fragility, the activities of the enzymes glucose-6-phosphate dehydrogenase (G6PD) and pyruvate-kinase (PK) and the concentrations of reduced glutathione (GSH) and 2,3DPG were evaluated in the erythrocytes. Diabetic dogs did not differ from controls in terms of haematological parameters, except for higher numbers of platelets. Higher values of fructosamine, glucose, protein, plasma potassium and calculated osmolality were detected in the plasma from diabetic dogs. No differences were detected in osmotic fragility, GSH concentration and PK activity between the two groups but 2,3DPG concentration and G6PD activity were statistically significantly higher in the diabetic group. The results indicate minimal alterations in erythrocyte functions occur in insulin-treated diabetic dogs. This contrasts with what has been reported for IDDM humans and cats. # 2002 Harcourt Publishers Ltd
INSULIN-DEPENDENT diabetes mellitus (IDDM) is one of the most common metabolic disorders occurring in dogs with a reported incidence as high as 15 per 1000 (Kaneko 1997). Haematological complications commonly develop in human diabetes mellitus. Erythrocyte morphology, functions and metabolism can be influenced by three general mechanisms: glycosylation of membrane proteins and enzymes, oxidative damage mainly due to glucose auto-oxidation, and changes related to ketoacidosis (Christopher 1995). These mechanisms cause haemoglobin glycosylation and altered oxygen affinity, decreased membrane deformability and alterations of metabolic patterns. They all contribute to shorten erythrocyte lifespan or to decreased oxygen availability to the tissue. Erythrocyte metabolism is mostly restricted to two main pathways. The Embden-Meyerhof pathway (EMP) is mainly involved in ATP and 2,3DPG generation in the erythrocytes, while the pentose phosphate pathway (PPP) generates NADPH, as a source of reducing equivalents to protect erythrocytes from oxidation. The activity of the enzymes pyruvate kinase (PK) and glucose-6-phosphate dehydrogenase (G6PD) are often tested as key-enzyme of EMP and PPP, respectively. Corresponding author: Dr Comazzi, Dipartimento di Patologia Animale, Igiene e SanitaÁ Pubblica Veterinaria, Sezione di Patologia Generale Veterinaria e Parassitologia, UniversitaÁ degli Studi, Via Celoria 10, 20133 Milan, Italy. Tel.: 0039-02-26680443; Fax: 0039-02-2364470; E-mail: [email protected]
0034-5288/02/010023 05 $35.00/0
In humans glycosylation affects the function of haemoglobin binding to 2,3-diphosphoglycerate (2,3DPG) site, thus influencing the regulation of the affinity with oxygen (Jones and Peterson 1981). However, in human medicine, findings concerning 2,3DPG concentration in the erythrocytes during diabetes mellitus are contradictory (Ditzel and Standl 1975, Kanter et al 1975). Decreases in G6PD activity are reported in human patients affected by complicated diabetes mellitus (Costagliola et al 1988, Costagliola 1990), contributing to increase erythrocyte susceptibility to oxidative stress. However, other authors have reported an increase in G6PD activity in uncontrolled diabetics (Goebel et al 1975). During diabetes mellitus many membrane changes due to oxidative stress are responsible for a decrease of osmotic resistance and of shortened erythrocyte lifespan (Gandhi and Chowdhury 1979, Jain 1989). Reduced glutathione (GSH) is one of the most important protecting agents by means of oxidation to a disulfide (GSSG). GSSG is reduced to GSH by glutathione reductase, using NADPH as a substrate to restore GSH concentration. Decreases in GSH concentration in diabetes mellitus have been reported both in human patients (Gandhi and Chowdhury 1979, EI Fakhri et al 1987, Konukoglu et al 1999) and in ketoacidotic cats (Christopher 1989, Christopher 1995). This has been associated with Heinz bodies formation (Christopher 1989). There are no reports regarding erythrocyte functions in canine diabetes mellitus. The haematological # 2002 Harcourt Publishers Ltd
24
S. Comazzi, S. Paltrinieri, V. Spagnolo, P. Sartorelli
complications during insulin therapy are very important because they can strongly influence both treatment and outcome in canine IDDM. In this study osmotic fragility, the activities of PK and G6PD and the concentrations of 2,3DPG and GSH have been investigated in order to better clarify the alterations of some erythrocyte functions in dogs affected by diabetes mellitus during insulin therapy. MATERIALS AND METHODS Blood samples were collected from 13 dogs affected by diabetes mellitus, at least 6 months from beginning insulin therapy, and from 36 healthy dogs that served as a control group. Insulin therapy had been chosen by the clinicians and administered once daily at the morning. Both well and poorly compensated dogs were chosen, but none of them showed clinical signs consistent with haematological or vascular complications. Five millilitres of blood were collected into lithium heparin tubes and 05 ml of blood was added to citratedextrose solution (ACD) anticoagulated tubes for GSH determination. The samples were immediately refrigerated at 4 C and all the tests on the erythrocytes were assayed within 2 hours. A complete blood cell count was performed by means of an automated analyser (H8 SEAC, Calenzano FI, Italy), and a differential leukocyte count was performed on a May-GruÈnwald Giemsa stained smear. Reticulocytes and Heinz bodies were evaluated on a brillant-cresyl blue stained smear (Pasquinelli 1984). Plasma was immediately separated and stored at ÿ30 C, and the samples were analysed for chemical assays within 2 weeks. Glucose (GOD-POD method, Hospitex Diagnostic, Firenze, Italy), total protein (biuret reaction, Hospitex Diagnostic, Firenze, Italy), urea nitrogen (enzymatic method, Hospitex Diagnostic, Firenze, Italy), albumin (bromocresol green colorimetric method, Hospitex Diagnostic, Firenze, Italy), inorganic phosphate (ammonium molybdate colorimetric assay, Hospitex Diagnostic, Firenze, Italy), fructosamine (nitroblue tetrazolium reduction, SEAC, Firenze) and b-hydroxybutyrate (Sigma Diagnostic, St Louis, US) were assayed on an automated analyser (Hospitex Diagnostic, Firenze, Italy) using commercial kits. Globulin was calculated subtracting albumin from total protein concentrations. Sodium and potassium were evaluated by a flame photometer (FP20, SEAC, Calenzano FI, Italy). Fibrinogen was estimated by means of a refractometer (Jain 1986). Osmolality was calculated as described by Carlson (1997). Osmotic fragility A 03 ml aliquot of blood was used to test osmotic fragility, according to the method of Dacie and Lewis (Miale 1982). Briefly, 25 ml of blood were mixed with 25 ml of the following concentrations of NaCl: 09 per cent; 075 per cent; 065 per cent; 055 per cent; 05 per cent; 045 per cent; 04 per cent; 035 per cent; 03 per cent; 02 per cent; 01 per cent. After an incubation
of 30 minutes at room temperature and centrifugation at 350 g for 10 minutes, the supernatants were read at a wavelength of 540 nm against water. 09 per cent NaCl was used as a negative control and the 01 per cent NaCl as a positive control from which complete haemolysis was obtained. From the values obtained, the fragiligram, composed of a cumulative (dose-response sigmoid curve) and a derivative curve (progressive haemolytic increment) of haemolysis, was drawn, and the concentrations of NaCl corresponding to the minimum (less than 5 per cent haemolysis), medium (50 per cent haemolysis) and maximum (more than 90 per cent haemolysis) osmotic fragility were calculated. 2,3DPG concentration An aliquot (05 ml) of blood was deproteinized in 15 ml of cold 8 per cent trichloroacetic acid. After centrifugation, the supernatant was stored at ÿ20 C and the 2,3DPG concentration was analysed within a month using a commercial kit (Sigma Diagnostic, St Louis, MO, USA), on a spectrophotometer at 340 nm. Haemoglobin concentration was tested by means of Drabkin's method (SEAC H10, Firenze, Italy) and the concentration of 2,3DPG was expressed in mmol gÿ1 Hb. Enzyme activities An aliquot (14 ml) of blood was centrifuged (15 minutes at 350 g) and plasma and buffy coat were carefully removed by aspiration. PK and G6PD activities were measured at 340 nm and 25 C in the haemolysate using commercially available kits (Boehringer Mannheim, Mannheim, Germany). Enzyme activities were expressed in mU gÿ1 Hb. Reduced glutathione (GSH) GSH concentration was evaluated spectrophotometrically at 412 nm, after deproteinisation with metaphosphoric acid (167 per cent) and EDTA (02 per cent) solution, using 5.5 0 -dithiobis-(2-nitrobenzoic acid) (Pasquinelli 1984). GSH concentration was expressed as mg/g Hb.
Statistical analysis Statistical analysis was performed by mean of a statistic software (Statistica, Stat Soft Inc, Tulsa, OK, USA). The t test for normally distributed data and the Mann-Whitney U test for nonparametric distributions were used. Correlations between haematology and chemical results and erythrocyte function were performed by the Spearman test. Results were considered significant when P < 005.
RESULTS The results of haematology and blood chemistry, which were significantly different between diabetic dogs
25
Erythrocytes metabolism in diabetic dogs TABLE 1: Haematology and blood chemistry results (mean and standard deviation), including number of cases and statistically significant differences between diabetic and control groups
Platelets n mlitre ÿ1 Total Protein g litre ÿ1 Globulin g litre ÿ1 Estimated fibrinogen mg dlitre ÿ1 Calculated osmolality mOsm litre ÿ1 Glucose mmol litre ÿ1 Fructosamine mmol litre ÿ1 Plasma potassium mmol litre ÿ1
Diabetic Mean (S.D.)
n
Controls Mean (S.D.)
n
P5
314917 (156877) 827 (117) 439 129 031 (021)
12
177077 (76586) 672 (78) 329 54 017 (013)
13
001
21
0001
21
001
17
005
12 12 10
341 (33)
11
316 (21)
13
005
1976 (1282) 434 (150) 481 (039)
12
538 (150) 291 (73) 438 (044)
13
0001
13
001
13
005
13 12
TABLE 2: Enzyme activities, 2,3DPG and GSH concentrations and osmotic fragility results (mean and standard deviation), number of cases and statistically significant differences between diabetic and control group; n.s. not significant
GSH /Hb
mg g ÿ1 Hb
PK
mU gÿ1 Hb G6PD mU gÿ1 Hb 23 DPG mmol g ÿ1 Hb min frag % NaCl med frag % NaCl max frag % NaCl
Diabetic Mean (S.D.)
n
Controls Mean (S.D.)
n
P5
234 (091) 161 (062) 491 (143) 1791 (465) 0486 (0060) 0414 (0040) 0322 (0061)
13
275 (098) 283 (217) 340 (120) 1437 (331) 0527 (0095) 0445 (0058) 0357 (0071)
13
n.s.
36
n.s.
36
0001
36
001
36
n.s.
36
n.s.
36
n.s.
13 13 12 13 13 13
DISCUSSION and the control group are shown in Table 1. No alterations in erythrocyte number, haemoglobin concentration, PCV, leukocyte counts and erythrocytic indexes were detected. Only higher platelet number was found (P < 001) in the diabetic dogs. No differences in the number of reticulocytes and nucleated RBCs were found, and no Heinz bodies were detectable in the blood from both groups. A slight lipaemia was noticed in some of serum samples even if no strong lipamic samples were found. Blood chemistry showed an increase of glucose and fructosamine in the diabetic group. Total protein and globulins were higher in the diabetic dogs as well as estimated fibrinogen without increases in albumin concentration. Calculated osmolality and plasma potassium were higher in the diabetic group. No alterations of b-hydroxybutyrate concentration were detected in the dogs examined. The results of osmotic fragility, enzymes activities and the concentrations of 2,3DPG and GSH are shown in Table 2. No statistical differences were detected in the activity of PK of the two groups but the mean value and the standard deviation were lower in the diabetic dogs while the distribution of the results in the control group was much wider. In contrast higher activities (P < 0001) of G6PD were detected in the diabetic group. No differences were found in GSH concentration but the mean value was lower in the diabetic group. In contrast diabetic dogs exhibited higher concentrations of 2,3DPG (P < 001) than control group. No differences were detectable in the osmotic fragility between the two groups. A positive correlation between plasma glucose concentrations and the activities of G6PD (P < 005, r 0664) and a negative correlation between the activities of G6PD and the concentrations of GSH in the erythrocytes (P < 005, r ÿ0626) were found.
The pathogenesis of canine diabetes mellitus strongly differs from feline and human maturity-onset forms and seems to be more similar to juvenile-onset diabetes in humans. The haematological complications of diabetes mellitus have been widely investigated in non-insulin dependent diabetes mellitus (NIDDM) of human beings (Jones and Peterson 1981) and partly in feline diabetes (Christopher 1989, Christopher 1995). However, no data are available about erythrocyte metabolism and functions in canine diabetes mellitus. Our results suggest few alterations of the haematological pattern occurs in insulin-treated diabetic dogs. Only platelets counts were higher in the diabetic dogs than in controls, but were still in the normal reference range. Anaemia has been reported more frequently in non-survivor diabetic dogs in comparison with survivor ones (Ling et al 1977), but this can sometimes be partly masked by dehydration associated with hyperglycaemia and hyperketonaemia (Christopher 1995). Anaemia was not found in any of our diabetic dogs and even if we cannot exclude dehydration as suggested by hyperosmolality in the diabetic group, was probably related to compensation induced by insulin therapy or to lesser numbers of damaged erythrocytes. Blood chemistry showed higher values of plasma glucose and fructosamine concentration in diabetic dogs, as was expected. Fructosamine is a very useful parameter to monitor the response to insulin treatment in diabetic dogs, because it reflects the average plasma glucose concentration over the preceding two weeks, and it is not affected by acute changes in the glucose concentration (Jensen 1995). In our dogs fructosamine concentrations fell into the range reported by Jensen (1995) for diabetic dogs (326±12429 mmol litreÿ1) and only one sample had a high concentration of fructosamine (788 mmol litreÿ1) while none of the dogs had ketosis. The glucose and fructosamine results showed a wide variation within the
26
S. Comazzi, S. Paltrinieri, V. Spagnolo, P. Sartorelli
population mainly because of one extreme outlying value. However, most values were higher than reference values indicating there was only partial regulation of diabetes in the group of dogs we examined. This presence of a high number of poorly controlled diabetic dogs together with the results of the correlation tests, suggest that the observed alterations of RBC metabolism are probably related to the severity of diabetes. Some evaluation in vitro to identify the direct effects of hyperglycaemia on enzyme activities and 2,3DPG concentration would be of interest. Plasma proteins, as well as fibrinogen, were higher in the diabetic group than in controls. This is due to an increase of globulins as no differences in albumin concentrations were detectable. Hyperfibrinogenaemia has been reported to be often related to diabetes mellitus both in human (Le Devehat 2000) and in veterinary species (Christopher 1995), however the magnitude of hyperfibrinogenemia in our samples is insufficient to account for the degree of hyperproteinemia. Moreover, even if no gross lipaemic samples were found in the present study it is possible some slightly lipaemic samples could have falsely raised the concentration of total protein, as lipaemia has been reported to interfere with some haematochemical and refractometer assays (Alleman 1990). A moderate increase in plasma potassium was detected in the blood collected from diabetic dogs with 5 out of 12 samples being higher than 50 mmol litreÿ1. Cotton et al (1971) reported hyperkalaemia in 29 per cent and hypokalaemia in 14 per cent of a sample of 28 diabetic dogs. In their sample, however, both well and poorly controlled diabetics were present and 11 of 28 dogs did not survive on insulin therapy. Hypokalaemia is often related to poorly managed insulin therapy and it is one of the most serious problems concerning insulin therapy and ketoacidosis (Nichols and Krenshaw 1995). When only considering the surviving dogs, the percentage of dogs with hyperkalaemia in Cotton's study increased to 39 per cent, which is similar to what we observed. The absence of hypokalaemia in the dogs we tested confirm the partial regulation of diabetes. In human diabetics an increase in osmotic fragility (Gandhi and Chowdhury 1979, Jain 1989) and a decrease in GSH concentration (Gandhi and Chowdhury 1979, Konukoglu et al 1999, Zaltzberg et al 1999) have been reported, and is often associated with diabetic complications (Thornalley et al 1996). Jain (1989) reported the increase of lipid peroxidation as one of the main causes of increased osmotic fragility of human erythrocyte in vitro. The dogs we observed did not exhibit any alteration of osmotic fragility or statistically significant decreases in GSH concentration. This could be related to a less damage to membrane structures in insulin-treated diabetic dogs, as suggested also by the absence of anaemic dogs in our samples. G6PD activity increased in diabetic dogs when measured in vitro. In vivo the activity of G6PD is regulated by many different conditions: changes in pH, the concentration of substrates (NADP and glucose-6 phosphate), inhibitors (nucleotides, metal ions, hormones, intermediates of glycolytic and polyol pathways) and activators (Ca, Mn, Mg). The activities of
G6PD in vitro were tested under optimal conditions, at stable pH and with excess NADP , but without other inhibitors or activators being added. Mature erythrocytes cannot synthesise new enzyme. Even though in vitro G6PD activity far exceeds the activity that occurs within the erythrocytes in vivo, we cannot exclude the possibility that in vitro increases of G6PD activities in diabetic dogs could reflect an in vivo upregulation of the enzyme in response to diabetic complications. The relation between this increase in activity and the severity of diabetes is suggested by the positive correlation between G6PD activities and glucose concentration and by the negative correlation between the G6PD activities and the concentration of GSH. Our results, obtained in insulin- treated patients, are partly in agreement with those obtained in humans, in which an increased activity of G6PD has been found in diabetic patients (Goebel et al 1975). On the other hand, other authors (Costagliola 1990) reported a decrease of G6PD activity in diabetic human patients, which might affect the integrity of the glutathione system. 2,3DPG concentration increased in the diabetic group. In previous studies, 2,3DPG concentration has been reported to increase in hypoxic circumstances due to cardiomyopathy, anaemias and polycythaemia vera (Paltrinieri et al 1998, Paltrinieri et al 2000, Comazzi et al 2000). In human affected by diabetes mellitus oxygen affinity has been reported to be abnormal, concomitant with the glycosylation of the 2,3DPG binding site or with some vascular diseases (Jones and Peterson 1981), and the concentration of 2,3DPG is reported to be markedly decreased in diabetic ketoacidosis (Ditzel and Standl 1975) and in poorly controlled diabetes (Triolo et al 1981). This is due to a lower plasma concentration of inorganic phosphate. In our samples, no alteration of inorganic phosphate were detectable. In well controlled patients or in patients with vascular complications, hypoxia is reported in association with an increase in 2,3DPG concentration thus decreasing the affinity of haemoglobin with oxygen (Ditzel and Standl 1975, Kanter et al 1975, Samaja et al 1982). In contrast, some authors (Marschner et al 1994, Marschner and Rietbrock 1994) have reported that 2,3DPG increases may have only slight effects on oxygen release in human diabetic patients. Canine glycosylated haemoglobin (Hb Can A0) has been described by Enoki and others (1986) with abnormal oxygenation properties, which is slightly different from the human form. In particular, a marked depression in the 2,3DPG and CO2 effects was reported and this could suggest a poor compensation to tissue hypoxia. However, according to previous studies in different hypoxic conditions (Paltrinieri et al 1998, Paltrinieri et al 2000, Comazzi et al 2000), we cannot exclude that the increases of 2,3DPG could represent a poor attempt to compensate for tissue hypoxia due to vascular complications and altered oxygenation properties of glycosylated haemoglobin in diabetic dogs. In conclusion the results of this study indicate minimal evidence of oxidative damage in erythrocytes from dogs affected by insulin-treated diabetes mellitus, compared to feline and human diabetes. In particular
Erythrocytes metabolism in diabetic dogs
no alterations in osmotic fragility nor Heinz bodies formation were found while the concentration of natural antioxidant, glutathione, is similar to healthy dogs. The evaluation of enzymatic activities in vitro showed higher activities of G6PD but, even if they showed correlation with glucose and GSH concentrations, more studies are needed to determine if they are directly related to diabetic complications. A higher concentration of 2,3DPG suggests a response to hypoxic condition that could be related both to haematological (altered affinity for oxygen) and non-haematological (vascular or thrombotic) complications. ACKNOWLEDGEMENTS Supported by MURST (ex 60 per cent). The authors wish to thank Dr Monica Merlini for technical assistance. REFERENCES ALLEMAN, A. R. (1990) The effects of hemolysis and lipemia on serum biochemical constituents. Veterinary Medicine 85, 1272±1284 CARLSON, G. P. (1997) Fluid, electrolyte and acid-base balance. In: Kaneko JJ, Harvey JW, Bruss ML eds. Clinical biochemistry of domestic animals. 5th ed. San Diego: Academic Press; 485±516 CHRISTOPHER, M. M. (1989) Relation of endogenous Heinz bodies to disease and anemia in cats: 120 cases. Journal of the American Veterinary Medical Association 194, 1089±1095 CHRISTOPHER, M. M. (1995) Hematologic complications of diabetes mellitus. Veterinary Clinics of North America: Small Animal Practice 25, 625±637 COMAZZI, S., PALTRINIERI, S., AGNES, F., SACCHET, A. & MILANI, F. (2000) Some aspects of erythrocyte metabolism in a dog with polycythaemia vera. Veterinary Record 147, 331±334 COSTAGLIOLA, C. (1990) Oxidative state of glutathione in red blood cells and plasma of diabetic patients: in vivo and in vitro study. Clinical Physiology & Biochemistry 8, 204±210 COSTAGLIOLA, C., IULIANO, G., MENZIONE, M., NESTI, A., SIMONELLI, F. & RINALDI, E. (1988) Systemic human disease as oxidative risk factors in cataractogenesis. Ophtalmic Research 20, 308±316 COTTON, R. B., CORNELIUS, L. M. & THERAN, P. (1971) Diabetes mellitus in the dog: a clinicopathologic study. Journal of the American Veterinary Medical Association 159, 863±870 DITZEL, J. & STANDL, E. (1975) The problem of tissue oxygenation in diabetes mellitus. Acta Medica Scandinavica 578 suppl., 59±69 EL FAKHRI, M., SHERIFF, D. S., CHANDRASENA, L., GUPTA, J. D., RAI, S. T. & QURESHI, M. S. (1987) Erythrocyte glutathione concentration in diabetics with cataracts, with and without glucose-6-phosphate dehydrogenase deficiency. Clinical Chemistry 33, 1936±1937 ENOKI, Y., OHGA, Y., SAKATA, S., KOHZUKI, H. & SHIMIZU, S. (1986) Structural and functional characterization of a canine minor glycosylated hemoglobin, Hb Can A1c. Hemoglobin 10, 607±621 GANDHI, C. R. & CHOWDHURY, D. R. (1979) Effect of diabetes mellitus on sialic acid and glutathione content of human erythrocytes of different ages. Indian Journal of Experimental Biology 17, 585±587 GOEBEL, K. M., GOEBEL, F. D., NEITZERT, A., HAUSMANN, L. & SCHNEIDER, J., (1975) Adaptation of red cell enzymes and intermediate in metabolic disorders. Enzyme 19, 201±211
27
JAIN, N. C. (1986) Schalm's veterinary hematology. 4th ed. Philadelphia, Pa: Lea & Febiger, 58±59 JAIN S. K. (1989) Hyperglycemia can cause membrane lipid peroxidation and osmotic fragility in human red blood cells. The Journal of Biological Chemistry 264, 21340±21345 JENSEN, A. L. (1995) Glycated blood proteins in canine diabetes mellitus. Veterinary Record 137, 401±405 JONES, R. L. & PETERSON, C. M. (1981) Hematologic Alterations in diabetes mellitus. American Journal of Medicine 70, 339±352 KANEKO, J. J. (1997) Carbohydrate metabolism and its disease. In: Kaneko, J.J., Harvey J. W. & Bruss M. L. eds. Clinical biochemistry of domestic animals. 5th ed. San Diego: Academic Press; 45±81 KANTER, Y, BESSMAN, S. P. & BESSMAN, A. (1975) Red cell 2,3-diphosphoglycerate levels among diabetic patients with and without vascular complications. Diabetes 24, 724±729 KONUKOGLU, D., AKCAY, T., DINCER, Y. & HATEMI, H. (1999) The susceptibility of red blood cells to autoxidation in type 2 diabetic patient with angiopathy. Metabolism 48, 205±299 LE DEVEHAT, C., KHODABANDEHLOU, T. & VIMEUX, M. (2000) Diabetes mellitus and fibrinogen. Hemorrheological and microcircolatory consequences. Journal des Maladie Vasculaires 25, 53±57 LING, G. V., LOWENSTINE, L. J., PULLEY, T. & KANEKO, J. J. (1977) Diabetes mellitus in dogs: a review of initial evaluation, immediate and long term management and outcome. Journal of the American Veterinary Medical Association 170, 521±530 MARSCHNER, J. P. & RIETBROCK, N. (1994) Oxygen release kinetics in healthy subjects and diabetic patients. I: the role of 2,3-diphosphoglycerate. International Journal of Clinical Pharmacology & Therapeutics 32, 533±535 MARSCHNER, J. P., SEIDLITZ, T. & RIETBROCK, N. (1994) Effect of 2,3diphosphoglycerate on O2-dissociation kinetics of hemoglobin and glycosylated hemoglobin using the stopped flow technique and an improved in vitro method for hemoglobin dissociation. International Journal of Clinical Pharmacology & Therapeutics 32, 116±121 MIALE, J. B. (1982) Laboratory medicine: hematology. 6th ed. St Louis: The C. V. Mosby Company, 883±884 NICHOLS, R. & CRENSHAW, K. L. (1995) Complications and concurrent disease associated with diabetic ketoacidosis and other severe forms of diabetes mellitus. Veterinary Clinics North American: Small Animal Practice 25, 617±624 PALTRINIERI, S., COMAZZI, S. & AGNES, F. (2000) Haematological parameters and altered erythrocyte metabolism in anaemic dogs. Journal of Comparative Pathology 122, 25±34 PALTRINIERI, S., SARTORELLI, P., DE VECCHI, B. & AGNES, F. (1998) Metabolic findings in the erythrocytes of cardiopathic and anaemic dogs. Journal of Comparative Pathology 118, 123±133 PASQUINELLI, F. (1984) Diagnostica e tecniche di laboratorio. Firenze, Italy: Rosini editore, 875±876 & 914±915 SAMAJA, M., MELOTTI, D., CARENINI, A. & POZZA G. (1982) Glycosylated haemoglobins and the oxygen affinity of whole blood. Diabetologia 23, 399±402 THORNALLEY, P. J., MCLELLAN, A. C., LO, T. W., BENN, J. & SONKSEN, P. H. (1996) Negative association between erythrocyte reduced glutathione concentration and diabetic complications. Clinical Science 91, 575±582 TRIOLO, G., GIARDINA, E., CONTORNO, A., CIPOLLA, T., GIANNUOLI, G., CAIMI, G., CATANIA, A. & RINALDI, A. et al. (1981) Glycosylated hemoglobins and erythrocyte 2,3-diphosphoglycerate in diabetes mellitus. Archivio per le Scienze Mediche 138, 179±185 ZALTZBERG, H., KANTER, Y., AVIRAM, M. & LEVY, Y. (1999) Increased plasma oxidizability and decreased erythrocyte and plasma antioxidative capacity in patients with NIDDM. Israel Medical Association Journal 1, 228±231
Accepted October 10, 2001