The Importance of Human Creatine Kinase Storage Conditions

Transcription

1 Ann Clin Biochem 1985; 22: 31~315 The stability of creatine kinase isoenzymes studied with two-site monoclonal antibody assays A P JACKSON and R J THOMPSON From the University of Cambridge, DepartmentofClinical Biochemistry, Addenbrooke's Hospital, Hills Road, Cambridge CB2 2QR, UK SUMMARY. The stability of human creatine kinase isoenzymes was investigated under different storage conditions using specific two-site monoclonal antibody assays. In Tris-HCI buffer ph 7 5 or barbitone buffer ph 8 1 containing 5 gil bovine serum albumin, the isoenzymes appeared to be stable for up to 3 weeks at 4 C but suffered a partial subunit dissociation and random reassociation after freeze-thawing; this dissociation was more pronounced as a result of freezing at -20 C rather than at -70 C. In contrast, creatine kinase isoenzymes stored in serum were stable at both 4 C and following freeze-thawing. High levels of heart type creatine kinase in serum showed only minor subunit hybridisation even after 12 h at room temperature. We conclude that in practical clinical situations, subunit hybridisation in serum samples is negligible. We recommend however, that isoenzyme standards for use in either two-site assays or radioimmunoassays should be stored frozen in normal serum from which endogenous creatine kinase isoenzymes have been previously removed. The dimeric enzyme creatine kinase (CK; EC ) exists as three tissue-specific isoenzymes: CK-MM (muscle-type), CK-BB (brain type) and CK-MB (heart type). Due to this restricted tissue distribution, CK-MB has been used as a diagnostic marker for myocardial infarctioni and CK-BB as a marker for head injury and some neurological disorders.? In order to interpret CK isoenzyme levels correctly, some information on their stability and the extent to which they undergo reversible subunit dissociation is required; this is particularly true if samples are stored for any length of time or are frozen and thawed prior to assay. There are several reports in the literature which suggest that CK isoenzymes are relatively labile'"? and can undergo significant hybridisation in vitro.": 4 Such studies have generally used enzyme activity measurements and there do not appear to be any similar reports using assays which measure mass concentrations. We recently developed two-site monoclonal antibody assays specific for CK-MB and CK BB. 6 We used these assays, together with a specific monoclonal antibody assay for CK MM, to re-investigate the problem of CK Correspondence: R J Thompson. 310 stability and subunit hybridisation in serum and commonly-used buffer systems under a variety of storage conditions. Materials and methods MATERIALS All chemicals were of Analar grade and purchased from BDH (Poole, Dorset, UK), except for Tris and bovine serum albumin which were purchased from Sigma (Poole, Dorset, UK). Cellulose ferrous oxide particles were the gift of Dr G Forrest, Serono Diagnostics Ltd. All radiochemicals were purchased from Amersham International (Amersham, Bucks, UK). CK-BB was purified from human brain as described previously' and CK-MM was purified from human skeletal muscle as described by Keutel et al.' CK-MB was purified from heart by a two-step immunoaffinity procedure using a low-affinity monoclonal antibody." All tissues were obtained within 12 h post mortem. TWO-SITE MONOCLONAL ANTIBODY ASSAYS Details of the CK-MB and CK-BB assays have been described elsewhere." The assay for CK BB employed an 12sI_labelled monoclonal anti-

2 Creatine kinase isoenzymes 311 body specific for CK-BB (which showed crossreactivity towards CK-MB of less than 0 1 % on a weight basis). The solid phase in this assay was a B-subunit-specific monoclonal antibody covalently bound to finely divided aminocellulose. Total incubation time was 5 h at room temperature; intra-assay variation was % between 0 1 and 1000 nglml; and inter-assay variation was <15%./1 The assay for CK-MB used an 125I-labelled B-subunit-specific monoclonal antibody with an immobilised M-subunitspecific antibody. In this assay the solid phase was a magnetisable cellulose-coated ferrous oxide suspension which allowed the rapid separation of bound label from free by using a permanent cobalt-samarian magnet.x Total assay time was 2 h at room temperature. The intra-assay variation was 14-19% between 0 5 and 1000 nglml whilst inter-assay variation was <21%./1 For the CK-MM assay a single M-subunitspecific monoclonal antibody (designated 15CH2) was used as both the label and solid phase. Iodinated 15CH2 was prepared by the Iodogen method" to give a specific activity of cpm/ng, whilst solid-phase antibody was prepared by coupling 15CH2 to cellulosecoated ferrous oxide particles by cyanogen bromide activation. III The resulting derivative contained 60 ug protein/mg solid phase and was stored as a 5 mglml suspension in 50 mm barbitone buffer (SSV) ph 8 1 containing 5 gil NaC!, 5 gil bovine serum albumin and 0 01 % (w/v) sodium azide. The CK-MM assay was performed as follows: magnetisable solid phase (0,05 ml) was added to 0 1 ml of either standard CK-MM or unknown in flat-bottomed test tubes held in a metal rack. After 1 h at room temperature on an orbital shaker, the tubes were washed with 1 ml of cold half-strength barbitone buffer containing 0 2% Triton X-IlK) and the solid phase rapidly precipitated using the magnetic separation system described by Forrest and Rattle. x The supernatants were decanted, the wash-step repeated and 0 1 ml of barbitone buffer containing \ cpm of iodinated 15CH2 was then added. After a further 2 h incubation with constant shaking at room temperature the washing and precipitation steps were repeated three times and the pellets counted for I min in an NE-1600 gamma counter. STORAGE OF CK ISOENZYME SAMPLES The effect of storage conditions on CK isoen- zymes was studied using the following buffer systems. (A) A 50 mm Tris-HCI buffer containing 10 mm 2-mercaptoethanol and 0 01 % sodium azide. (B) Sodium barbitone buffer with the composition described above. (C) Pooled normal serum containing 0 01 % sodium azide from which endogenous CK-MM and CK-MB had been removed by the following procedure. Serum (25 ml) was incubated at 4 C for 24 h with monoclonal antibody 15CH2 covalently bound to finely divided aminocellulose (10 mg of cellulose containing approximately 1 mg protein)." The serum was then centrifuged at 2400 rpm for 10 min and the supernatant retained. No CK isoenzymes were detectable in this supernatant when measured with the two-site assays. Purified CK isoenzymes were added to each of these solutions in the following combinations: (1) A high CK-MB level (700 nglml). (2) A low CK-MB level (2 nglml). (3) A high CK-MM level (600 ng/rnl.) in combination with a high CK-BB level (30' nglml). (4) A high CK-MM level (600 ng/rnl.) in combination with a low CK-BB level (1 5 nglml). Each of these combinations in serum, barbitone and Tris buffer were subdivided into three aliquots and the levels of CK isoenzymes measured with the two-site assays. One aliquot of each combination was then frozen at -70 C, a second aliquot of each at - 20 C and the third left at 4 C. Each CK isoenzyme was then re-measured in all the samples between 2 and 3 weeks after the start of the experiment. Quality-control samples containing a high and a low isoenzyme level were included in all assays performed. These quality-control samples were stored in pooled normal serum at - 70 C until used. Each individual isoenzyme was also measured in serum samples from two patients with confirmed myocardial infarction. Aliquots of these sera were then stored at -70 C, at - 20 C and at 4 C and the isoenzymes re-measured up to I week later in order to confirm that the results obtained with sera also applied to genuine clinical samples. Finally, sera initially containing a high (700 nglml) and a low (2 nglml) level of CK-MB were left at room temperature for 12 h to investigate whether changes in isoenzyme com-

3 312 Jackson and Thompson FIG. 1. Standard curve for CK-MM assay. Total counts added= 100 OOO/min. Points are means±sd of four determinations. Also shown is the effect of ng/ml CK-BB (A) and 5000 ng/ml CK-MB (0) at O. 10 and 500 ng/ml CK-MM. 100 'd'i-j':l.- ---,l I -l.. I..JI CK-MM Ing/mLl position are likely to occur during collection and short-term storage. Results CK-MM ASSAY Fig. 1 shows a standard curve obtained with the two-site monoclonal antibody assay for CK MM. There was no interference from CK-BB up to at least nglml and none from CK-MB up top 5000 ng/ml (Fig. 1). The intra-assay variation was % between the limits of 5 and 100 nglml. Inter-assay variation was 10% at 500 nglml. Standard curves for the CK-MB and CK-BB assay have been shown previously." EFFECT OF STORAGE CONDITIONS ON CK ISOENZYMES The results of the storage and freeze-thaw experiments are presented in Tables 1-5. The levels of all three isoenzyrnes were measured at the start of the experiment and after storage for the stated periods at 4 C, -20 C and -70 C in serum, Tris-HCl and sodium barbitone buffer. Changes accompanying the storage of CK-MB at room temperature are shown also. Values are means of duplicates agreeing to within 5%. The levels of immunoreactive CK isoenzymes in all three systems remained stable at 4 C with no evidence of subunit hybridisation. Moreover, CK isoenzymes stored in serum showed TABLE 1. CK isoenzyme levels after storage in pooled normal serum Sample no Storage Initial Final level (ng/ml) following storage at: period level (weeks) Isoenzyme (ng/ml) 4 C -20 C -70 C 2 CK-MM 0 0 () () CK-MB 7lJlJ CK-BB () CK-MB CK-BB () 3 CK-MM Xl CK-MB CK-BB /l 26 3 CK-MM 5lJlJ lKJ CK-MB CK-BB \ 3 \ 3 \ 3 \ 4

4 Creatine kinase isoenzymes 313 no significant change in mass concentration and no evidence of subunit hybridisation following freeze-thawing at either - 20 C or at -70 C (Table 1). This stability to freeze-thawing is also shown by serum samples from myocardial infarct patients containing raised CK-MB and raised CK-MM (Table 4). When frozen and thawed in either Tris-HCl buffer or barbitone buffer there were significant changes in isoenzyme levels: CK-MB appeared in samples originally containing only CK-MM and CK-BB and similarly, CK-BB appeared in samples originally containing only CK-MB (Tables 2 and 3). This partial subunit dissociation and random re-association was far more prominent following freezing at - 20 C compared with -70 C (Tables 2 and 3). It was however not related to the time spent frozen TABLE 2. CK isoenzyme levels after storage in sodium barbitone buffer Sample no Storage Initial Final level (ng/ml) following storage at: period level (weeks) Isoenzyme (ng/ml) 4 C -20 C -70 C CK-MB CK-BB CK-MB CK-BB CK-MM CK-MB CK-BB CK-MM CK-MB CK-BB TABLE 3. CK isoenzyme levels after storage in 50 mx Tris-HCI buffer Storage Initial Final level (ng/ml) following storage at: Sample period level no. (weeks) Isoenzyme (ng/ml) 4 C -20 C -70 C CK-MB CK-BB CK-MB 2 2 I 1 1 CK-BB CK-MM 600 5SO CK-MB I CK-BB IS CK-MM CK-MB CK-BB TABLE 4. Changes in CK isoenzyme levels in serum from myocardial infarct patients following storage Initial Final level (ng/ml) following storage at: Sample Storage level no. period Isoenzyme (ng/ml) 4 C -20 C -70 C I week CK-MM SOO 790 SOO 800 CK-MB S(J CK-BB days CK-MM CK-MB (J CK-BB «n

5 314 Jackson and Thompson TABLE 5. Changes in CK isoenzyme following storage at room temperature Sample Storage Initial level Final level no. period Isoenzyme (ng/ml) (ng/ml) 12 h CK-MM 0 () CK-MB CK-BB () h CK-MM () 0 CK-MB 3 2 CK-BB U () since the dissociation of CK-MB in Tris-HCI buffer was no worse after 3 weeks at - 20 C than it was after 1 day at -20 C (data not shown). The loss of CK-MB, where this occurs (Tables 2 and 3) was also generally greater than the gain in CK-BB or CK-MM, indicating that the subunits underwent not only a partial dissociation but also a partial denaturation as well. Some evidence for limited subunit hybridisation was found with high levels of CK-MB in serum at room temperature (Table 5) but this occurred only to a minor degree. Discussion Previous reports on the stability of CK using activity measurements have found that all three isoenzymes are labile unless stored with a variety of thiol compounds.y" In contrast, our results for CK in serum suggest that if mass concentration is used rather than enzyme activity, the isoenzymes are quite stable both at 4 C and following freeze-thawing. The practical consequence is that immunological assays measuring mass concentration and which are generally more sensitive than enzyme activity measurements" should also be more reproducible when used on stored samples. We have also investigated the extent to which subunit hybridisation occurs between creatine kinase isoenzymes in solution. There are at least two major experimental approaches which can be used to study this problem. In one method the degree of subunit interchange may be followed by incubating one radioactively labelled isoenzyme in the presence of a second unlabelled isoenzyme (e.g. 125I-labelled CK-BB incubated with unlabelled CK-MM). Subunit hybridisation could then be monitored by measuring the radioactivity in all three isoenzymes after a defined period. It is however, possible that inter-subunit stability may itself be affected by the labelling procedure used. An alternative approach is to monitor changes in the levels of unlabelled proteins using isoenzyme-specific assays. This is the method we have adopted using the specific monoclonal antibody assays and it is also the method employed in previous studies which have used cellulose acetate electrophoresis and activity staining to quantify each isoenzyrner'' " In contrast with these earlier studies": 4 we find no evidence of subunit hybridisation in unfrozen samples at 4 C or in frozen and thawed serum (Tables 1--4). We also find only minor dissociation after 12 h at room temperature (Table 5). The reasons for this discrepancy are unclear and may require further investigation. A significant dissociation and partial reassociation did occur when CK isoenzymes were frozen and thawed in either barbitone or Tris-HCI buffer (Tables 2 and 3). The former buffer contained 5 gil bovine serum albumin which is known to protect proteins against desaturation. II The subunit hybridisation of lactate dehydrogenase after freeze-thawing is nevertheless not prevented by up to 10 gil albumin.f It is therefore possible that protection against hybridisation only occurs with extremely high protein concentrations as found, for example, in serum (Tables 1 and 4). It is also noteworthy that subunit hybridisation was more pronounced following freezing at -20 C compared with -7()OC (Tables 2 and 3). Several factors may be responsible. Firstly, freezing a buffer solution to a temperature above its eutectic point will produce local regions of unfrozen liquid containing high concentrations of salt and acid which may then cause dissociation and denaturation." Secondly, a low temperature itself is known to weaken hydrophobic interactions between the subunits of oligomeric proteins13 and finally, differences in the rates of cooling at the two temperatures may play a role." These three

6 Creatine kinase isoenzymes 315 factors may combine to give the observed differences in temperature stability. Until recently, we have stored our CK standards for use in radioimmunoassays at -70 C in a 50 mm Tris HCl buffer containing 5 g/l bovine serum albumin. However, storage of CK-MB standards in this manner led to the appearance of some 1-2'Yo contamination with CK-BB as measured with the two-site monoclonal CK-BB assay. In view of our results we propose in future to store our CK standards in pooled normal serum from which endogenous CK has been removed. The most straightforward way to achieve this is by batch adsorption with appropriate monoclonal antibodies covalently bound to a suitable solid phase. Acknowledgements This work was supported by a grant from the East Anglian Health Authority to RJT. APJ holds a Medical Research Council Studentship for training in research methods. References Roberts R, Gowda KS, Ludbrook PA, et al. Specificity of serum MB creatine phosphokinase activity in the diagnosis of acute myocardial infarction. Am J Cardiol 1975; 36: 43J-7. 2 Thompson RJ. Graham JG. McOueen JNF. et al. Radioimmunoassay of brain type creatine kinase BB isoenzyme in human tissue and sera of patients with neurological disorders J Neural Sci 1980; 47: Morin LG. Creatine kinase: stability, inactivation. reactivation. cu«chem 1977; 23: Szasz G. Gerhardt W, Gruber W. Creatine kinase in serum: 5. Effects of thiols on isoenzyme activity during storage at various temperatures. Clin Chem 1978; 24: {l3. 5 Nealon DA, Henderson AR. Stability of commonly used thiols and of human creatine kinase, isoenzymes during storage at various temperatures in various media. Clin Chem 1977; 23: Jackson AP, Siddle K. Thompson RJ. Two site monoclonal antibody assays for human heart and brain-type creatine kinase. Clin Chem 1984; 30: I I 57---{l2. 7 Keutel J, Okabe K. Jacobs HK, et al. Studies on adenosine-triphosphate-creatine transphosphorylase XI. Arch Biochem Biophys 1972; 150: Forest GC, Rattle SJ. Magnetic particle radioimmunoassay. In: Hunter WM. Corrie JET. eds. Immunoassays for Clinical Chemistry, 2nd edn. Edinburgh: Churchill Livingstone, 1983; {l2. 9 Fraker Pl, Speck lc. Protein and cell membrane iodinates with a sparingly soluble chloroamide 1,3.4,6-tetrachloro-3a, 6a-diphenylglycol uri!. Biochim Biophys Res Commun 1978; 80: Axen R, Porath J, Ernback S. Chemical coupling of peptides and proteins to polysacharides by means of cyanogen halides. Nature (London) 1967; 214: II Shikama K, Yamazaki J. Denaturing of catalase by freezing and thawing. Nature (London) 1961; 190: Chilson OP, Costello LA. Kaplan NO. Effects of freezing on enzymes. Fed Proc 24 (Supp!. 15): S55-S Gutfreund H. In: Wiley J. cd. Enzymes physical principles, Accepted for publication 13 September 1984