Ann. Hum. Genet., Lond. (1979), 43, 15
15
Printed in Great Britain
The expression of creatine kinase isozymes in human cultured cells BY SUSAN POVEY," MARY INWOOD," ALISON TANYAR" AND MARTIN BOBROW?
* MRC Human Biochemical Genetics Unit, Galton Laboratory, University College London and Genetics Laboratory, Department of Biochemistry, Oxford INTRODUCTION
Three isozymes of creatine kinase have been identified in human tissues, the MM form characteristic of skeletal muscle, the BB form which is predominant in brain, and the heteromeric MB form which is most active in heart muscle (Burger, Richterich & Aebi, 1964; Dawson, Eppenberger & Kaplan, 1965). Sensitive immunological techniques and quantitative assay after ion exchange chromatography have detected varying proportions of these isozymes in other tissues (Wretou & Pfleiderer, 1975; Tsung, 1976). It has been reported that the BB isozyme occurs in human foetal fibroblasts (Maienhofer et al. 1970) and also that the MM form may occur in lymphocytes and polymorphonuclear leucocytes (Cho et al. 1977). We have investigated the occurrence of creatine kinase in human lymphoid lines, fibroblasts and also rodent cell lines and human-rodent somatic cell hybrids.
MATERIALS AND METHODS
Human tissues, usually brain, heart and skeletal muscle, were available in the course of other investigations and were stored at - 20 "C. Mouse and hamster tissues were stored in the same way. Human fibroblasts were cultured from small skin biopsies using routine methods, and were maintained in Eagle's MEM with 10 yo foetal calf serum. Human lymphoblastoid lines, which were originally kindly supplied by Dr Michael Steel, were maintained in RPMI 1640 with 20 yo foetal calf serum. Rodent cells and human-rodent somatic cell hybrids were supplied to us in the course of other projects by various laboratories (see Table 1) and were transported as frozen cell pellets. The origins of the various rodent. cell lines tested are given in the references to Table 1, except for the hamster line V79 (Chu et a&.1976). Cultured fibroblasts were examined where possible on the day of harvesting. Other cultured cells were usually stored in liquid nitrogen until they could be analysed. Tissues were extracted by homogenization in distilled water ( l : l ,w/v) using a Silverson emulsifier. Washed pellets of lymphoid lines (approx. 5 x lo7 cells) were mixed with a volume of water approximately equal to the cell pellet and subjected to two short bursts of sonication from a probe sonicator, the sample being surrounded by ice. Homogenates and sonicates were then spun at 12 000 g for 20 min and the supernatant used for the investigation of CK. For the cultured fibroblasts a pellet of approximately 4 x lo6 cells was sonicated in the smallest possible volume of water and the whole sample applied to the gel. Rodent cells and human-rodent hybrids were usually extracted as for lymphoid lines. For the investigation of CK in the various human cell lines starch gel electrophoresis was 0003-4800/79/0000-4304 $03.00 @ 1979 University College London
16
SUSAN POVEY AND
OTHERS
Fig. 1. Starch gel showing creatine kinase BB isozymes in a number of different lymplioid lines. Approximately the same amount of sample from each line was applied to the gel. Elcctroplioresis hist,idine-citrate,p H 7.0. Tetrazolium-linked stain.
carried out using 0.005 M histidine/NaOH p H 7.0 as the gel buffer and 0.41 BI citric acid pH 7.0 as bridge buffer (Harris 8: Hopkinson, 1976). The gel was run a t 2 V/cm for 18 h with cooling. On occasion an alternative buffer system, Tris-citrate pH 7.5, was used; bridge 0.094 11 Tris/0.0235 M citrate, gel a 1 in 5 dilution of this; running conditions as before. For the separation of the human and rodent BB isozymes a citrate phosphate buffer system pH 5.8 n-as used (bridge 0.11 M citrate, 0,245 M phosphate, gel a 1 in 40 dilution of this). The gel was run at 4 V/cm for 17 h with cooling. Detection of creatine kinase was by two standard methods, either using creatine and a negative fluorescent stain, or using creatine phosphate and a tetrazolium-linked stain (Harris 8: Hopkinson, 1976). When the tetrazolium stain was used the concentration of creatine phosphate was increased approximately two-fold over that described. I n general, activity in the cultured cells was rather low and the gels were incubated with the stain for 1-2 h a t 37 "C. Human chromosome identification in the hybrids was by G-11 staining followed by Q-banding as described by Bobrow 8: Cross (1974). RESULTS
Human cells Fibroblasts from 20 individuals (10 adults, 9 foetuses and 1 amniotic cell culture) together with 46 lymphoid lines from 42 individuals were screened for the presence of' CK isozymes. The lymphoid lines included those from 12 patients with malignant disease (mostly Burlcitt lymphomas or leukaemias), from 16 patients with a wide range of other disorders (including infectious mononucleosis, cystic fibrosis and X chromosome abnormalities) and from 14 healthy babies and adults.
C’reatine kinase isozyrnes in human cells
17
Fig. 2. Opposite halves of a starch gel stained with and without creatine in t,he negative fluorescent stain. ( a ) Non-specific isozymes; ( b ) shows in addition true creatine kiriase isozynies in heart, brain and the lymphoid line TAY,. Electrophoresis histidine-citrate, pH 7.0.
Thc BB isoxyme was detected satisfactorily by either staining method, and the bands observed were always dependent on the presence of creatine or creatine phosphate in the stain. In all fibroblasts tested the BB isozyme was present and no consistent differences between adults and foetuses were observed. Seventeen of the lymphoid lines from 16 different individuals mere clearly positive for the CK BB isozyme, but there was considerable variation in the amount of activity expressed (Fig. 1 ) . For example, three separate lines T A P 2, TAY 3 and TAU 4 were started on the same day from the same individual, a patient with acute myeloid leukaemia (Evans, Steel & Arthur, 1974). TAY 4 consistently had very high creatine kinase activity, similar to that of brain, whereas in TAY 2 the activity was much less and in TAY3 it was not detected. CK BB was found in 8 out of the 16 lines from healthy individuals (including both adults and babies), in 3 out of 14 lines from patients with malignancy (including TAY 2 and TAY 4) and in 6 out of 16 lines from patients with other diseases. He\ era1 other enzymes including alkaline phosphatase and malic enzyme show considerable line-to-line variation in activity, but neither of these latter enzymes showed a definite correlation with the expression of CK. However, the lines negative for malic enzyme (exclusively lymphomas and lymphatic leukaemias) have so far also been negative for CK BB and will be reported in detail elsewhere. The M B and M M isozymes were generally investigated using the negative fluorescent stain, since the tetrazolium-linked stain has the disadvantage that adenylate kinase (AK)isozymes also appear on the gel. In the fibroblasts the results were clear and no MB or MM isozymes were observed. Much more difficulty was encountered in the assessment of the lymphoid lines. Varying amounts of components in the MM and MB position were seen in the lines, but these 2
H G E 43
SCSAX POVEY AND OTHERS
did not clep,rntl on tlie presence of creatine in the stain (Fig. 3 ) . Similar results were obtained in all th13t.t~elcctroplioretic systems. The bands in the JIB position depend on the presence of both A'I'P ;iad PEP (pliosl)lioenolp?-rurate)in the stain. They may be non-specific phosphatases, but iond in position to nnj- of the well characterized human phosphstases (Harris & Flopkinson. 1 O i f i ) . The isozj-mes soiuetimes seen in the At31 position are reduced in iwtivity by high \ l ) t ' t ~ lcrnti*ifiigation(100000 g for 1 11) of tlie extracts and also do not require PEP for actij-itj I t ~t'euispo-sible that this actirity is attributable to membrane-bound hydrolases, althotigh the n-aj in which they are revealed by this stain is not clear. Bec,iiiw of tlic tlifficulties with the negative stain the gels were also stained using the tetrazoliuiii-linkctl staiti. Xi1 at tempt was madr. to inhibit the AK isozymes using adenosine monophoy)Ii~ite(tip to 1 mgZ/ni1in stain) as suggested by Rosalki (1967). b u t this was only partially succes.f~il. F,tiiit crea tine lciiiase JIM isozyines could have been obscured by AK isozynies (altho1igl~-lK1 i4 bai-clj detectable in most lymplioicl lines), but the JIB isozyme should have i l k . So ti ~ C of C JIB or JZJI isozyne was seen in any lymphoid line tested.
Creatine Einase isozymes in human cells
19
Fig. 4. Part of a starch gel showing creatine kinase isozymes in hybrids made with the mouse line P G 19. The insert line of the gel is not shown. Electrophoresis citrate-phosphate pH 5.8. Tetrazoliumlinked stain.
Rodent cells
A survey of mouse and Syrian and Chinese hamster tissues, using both stains, revealed a distribution of isozymes similarto that of the human. After electrophoresis in citrate phosphate pH 5.8 the human and mouse BB isozymes were clearly distinguishable, the human being more anodal than either hamster or mouse (Fig. 3). I n view of the findings in the human cultured cells, the study of the cultured rodent cells was centred on the BB isozymes. For this the tetrazolium-linked stain was used as it appeared to be slightly more sensitive and certainly gave sharper bands. The mouse line PG 19, originally derived from a melanoma (Jonasson, Povey & Harris, 1977) had a clear band of CK BB in the same position as the brain enzyme (Fig. 3). The other mouse lines tested, lR, 3T3, A9HT and RAG and the hamster lines V79 and WgSh, appeared to lack creatine kinase. Hybrids
A total of 31 primary independent hybrids were tested for CK. Hybrids made between human cells and PG 19 were clearly scorable for the presence or absence of human CK B polypeptide. Positive hybrids had a three-banded pattern consistent with the dimeric nature of CK BB, whereas negative hybrids had the mouse enzyme only (Fig. 4).Hybrids made with other rodent lines either had no CK or the human CK BB enzyme only (Fig. 5). The right-hand column of Table 1 indicates in which crosses human CK BB was seen in at least one clone. Analysis In the analysis of a differentiated function four classes of human-rodent hybrid can be envisaged, i.e. both parental cells expressing the function, neither expressing it, or expression occurring in the human or rodent parent only. We have studied hybrids in all four groups and 2-2
20
SUSAN POVEY AXD OTHERS
Fig. 5 . Starch gel showing creatine kinase isozymes in three different mouse lines and human-rodent hybrids made with them, showing that human CK BB can be expressed independently of the mouse enzyme. Tetrazolium-linked stain. Electrophoresis citrate-phosphate, p H 5.8. (PG 19 has a characteristic two-banded pattern for AK, and may be a heterozygote).
also some in which the status of the human parent with regard to CK BB was unknown (Table 1). From our results so far it appears that the mouse enzyme in PG 19 is not extinguished by fusion with human cells, and is uninfluenced by human chromosomes. Similarly, the lack of CK BB in the other rodent lines was not affected by fusion. The presence of human CK alone in some of the hybrids made with these cells provides good confirmation that these rodent lines are truly not expressing CK BB and the failure to demonstrate it is not due to technical reasons. I n the few hybrids made with human parental lines known not to express CK (the lymphoid lines MOLT, and RAJI) the human CK BB isozyme has not been seen, but this cannot exclude the possibility that reactivation of human CK BB might occur in such a hybrid. Partly because of the negative results, but also because both RAJI and MOLT, are highly aneuploid these hybrids have been excluded from the subsequent analysis. I n an ideal situation one would analyse each cross separately, but insufficient data are available to do this usefully a t present. Table 2 therefore shows a summary of the segregation of human CK BB in all 28 independent hybrids (excluding those made with RAJI and MOLT,). There is some discordance with all chromosomes. This is not unexpected since on a background where the mouse enzyme is not expressed complex factors may determine the expression of the human gene. Even so, it seems reasonable to assume that human CK BB cannot be expressed in
25
Normal fibroblasts X/I 5 translocation fibroblast I / I I translocation fibroblast I /17 translocation fibroblast Normal fibroblast Normal fibroblast Lesch Nyhan fibroblasts MOLT, lymphoid RAJI lymphoid T lymphocytes BRE lymphocytes
Fisher el al. '977 Solomon et al. '976 Povey et al. I976 Povey et al. I976 Solomon et al. I979 F'ellous et al. '973 Carritt et al. '977 Unpublished (Oxford) Dr Rosalie Ber Jones et al. 1976 Sant,achiara et al. 1970 van Heyningen et al. 1975 Lymphocytes
Human parent
Reference or source
4 I I
I Rmouse
0
2 I0
I
8
0
AgHT mouse P G I 9 mouse 3T3 mouse
2
20
-
A 23 hamster PG 19 mouse
0
I
LMTK- mouse
2
8
3
I
8
2
0
RAG mouse
A23 hamster
5
I
I Rmouse Wg3h hamster
7
p G 1 9 mouse
Rodent parent
No
No Yes Yes
No
Yes
N O
Yes
Yes
Yes
Yes
Yes
Human CK BB seen in hybrid
Parentheses indicate that, although CK was not tested on the parental cells, it was assumed t o be present as no example of human fibroblasts not expressing CK BB has been found.
HORP
AgHT/RAJI TP Z412
MOP
PF
P7A/2
MOG
8z/Az3
81 /wg
DURq
DFP
Name of hybrid
No. of indepen- No. of subdent hybrids clones
Table 1. Expression of human CK BB in human-rodent hybrids
22
SUSAN POVEY AND OTHERS
Table 2. Segregation of human creatine kinase BB and chromosomes as assessed by marker enzymes and/or karyotyping in 28 independent primary hybrids CK/chromosomes ,-----Jc------
Chromosome I 2
3*
4 5
6 7 8* 9 10 I1
I2
I3
14
PGM,/ENO,/PGD/PEPC MDHJACP, GPX PGM, HEX, ME,/PGM,/GLO JIGUS GSR AK,/AK,/ACON,
GOT,
8 8 4
9
5 '4 7 5 9
8 I2
ESD NP
5 16
19 20
ADA
16 I7 I8
21
SOD,
X
G6PD
-/-
I4 7
I0
LDH, LDH,/PEPB MPI/PK APRT GALK PEPA GPI
15
+/+
Enzymes
9
I5 I1
12 15
7
+/-
-/+
5
4
2
3
I
2
3
I
4
5 4 5
6
I
2
8
4
3
2
2 2
7 8 7 7
5 5 7 4 6
4
7 7
0
5 5
4
5 4 7 4 6
5
5
6
9
15
5 5
I 0 I
I0 I
I
5 4 I
3
7 7
9 6 '4 I 2 I1 * Enzyme markers used in only a few hybrids - most of the assessments of these two chromosomes (3 and 8)were from karyotype analysis.
the absence of its structural gene. There are two chromosomes (14 and 1 7 ) which are present in every hybrid expressing human CK, and a further five chromosomes which have been recorded as absent in the presence of CK BB on only one occasion (3, 4, 15, 18, 20). Analysis of subclones from five independent positive hybrids was therefore undertaken. Segregation of CK BB occurred in three of the hybrids. Details of the chromosomes in subclones of two of these are shown in Table 3 (the others were not karyotyped). Definite examples of the retention of CK BB in the absence of chromosomes 3, 4, 15, 18 and 20 were found. The segregation of human CK BB with chromosomes 14 and 17 in subclones of all five hybrids is summarized in Table 4. Three further sets of subclones were tested for CK BB, although the original human-hamster hybrids (PF, see Table 1) had not been tested. The segregation of human CK BB and chromosome 14 in those subclones is also shown in Table 4. These hybrids were assumed to have chromosome 17 because they were originally made by selection for thymidine kinase and maintained in HAT medium. DISCUSSION
If a differentiated function is defined as a gene product found very much more in some types of cells in a living organism than others, creatine kinase BB must certainly be regarded as such. The simplest explanation of the striking differences in creatine kinase BB activity in the cultured cells found here is that this is a reflexion of variation of CK BB activity in their various cells of origin. Although it cannot be excluded that some of the lines may have lost
+
-
DTI-5
-
-
-
-
-
6
-
rt
7
+
-
-
-
10
+
i
11
+ - + -
-
+ -
9
8
.
.
-
+ - .
.
-
-
+ -
*
-
+ + + -
+ + ++ + + +
g
ZI
22
+ + +
20
-
X 74+
- - - + + f + -+/?+/-- + - + - - - + + - - + + - * + - + + + + + - - + + + . + . . +
2
16 17 18 19
++/-+/-+
14 1 5
+ - -
f
-
-
-
13
+ + + + + + + + + + + . . + + + + - +
*
-
+
IZ
of cells; t , chromosome seen in 30 cliromosomo uncertain ; + / - , discrepancy (ciixynie/chroniosorne).
+ , Chromosome seen in >
DTI-5 which was not karyotyped.
< 30 76 of cells; L - , cllromosornc seen in occasional cell;
?, identification of
The presence of human chromosomes was assessed primarily hy lraryotyping but checked by enzyme studies in most cases. The exception was
-
+
+
-
5
- - - - - - - +/+ - - + - - - - + - + - - - -
+
-
+ + -
4
+
3
&
+
2
- + + + - + - - +
- -
-
-
-
+
Z
+
18 DURq DUR4R
I2
I0
Z412 Clones of Z412 4
I
CK
Table 3. Xegregation of CK BB and human chromosomes in subclones of two independent hybrids
3
p.3
w
3
s-
cn
3 m
*.
2
Q
w *.
h
s2.
cl
24
SUSAN POVEY AND OTHERS
Table 4.Segregation of CK BB with NPlchromosome 14 ant1 GALKlchromosome 17 in all subclones tested
-I+ Subclones from five NP/ I4 positive hybrids (pooled) GALK/I7 Subclones from three untested hybrids (a) NP/w GA4LK/I 7
(b) (c)
2
16
0
0
7 7
0 0
0 0 0
NP/ I4 GALK/I 7
I
5
I
0
0
SP/I4 G=~LK/I~
I
0
0
I
0
0
7
CK BB activity as a result of long-term culture, it seems possible that the study of CK BB in hybrids may shed some light on the process of differentiation. Snalysis of human-rodent hybrids described here has suggested that the presence of either chromosome 14 or 17 or possibly both, is necessary for the expression of human CK BB, and there is some evidence that CK BB activity is lost if chromosome 14 is lost. Even in thc cross where both parents are positive for CK BB neither 14 nor 1 7 alone is sufficient for the expression of CK BB. There are also two hybrids related t o the positive JIOG3.6A which have enzyme markers for both 14 and 1 7 and yet do not express CK BB. The simplest explanation of these results is that the structural locus CK, is on chromosome 14 and that other human chromosomes are necessary for its expression. So far it has not been possible to identify the specific chroniosomes involved - this may be because of insufficient data, or because different chromosomes are involved in different lines. Another possible explanation is that we are not looking a t the cffect of particular human chromosomes, but of some factor(s) related to the human contribution whilst not absolutely related to the human chromosomes. It may be relevant that the only 31OG hybrid to express human CK BB appears to have virtually a full set of human chroniosomes. Clearly more data are needed. It is a feature of most human-rodent hybrids that the rodent karyotype remains almost complete whereas the human contribution varies. I n the case of CK BB e~aetlgrthe same o c ( ~ r s , the rodent phenotype being constant and the human variable. This suggests either that diffusiblc ‘activators’ or ‘repressors’ do not exist, or that they are species-specific. It would therefore br useful to look a t some intra-specific hybrids, but the value of these would be greatly increased if a variant of CK BB were used, since the different parental contributions could be assessed separately. Unfortunately, no variants of CK BB have yet been reported.
SUMMARY
The BB isozyme of creatine kinase is consistently present in cultured human fibroblasts and shows great variation in activity in long-term lymphoid lines. One mouse line tested, PG 19, had strong activity, but all other rodent lines tested did not express CK BB. Human and mouse CK BB can be expressed independently of each other in human-rodent somatic cell hybrids.
Creatine kinase isozymes in human cells
25
There is some evidence that the structural locus for CK BB may be on chromosome 14, but the involvement of other chromosomes, especially no. 17, cannot be excluded. The MM and MB isozymes of creatine kinase were not seen in any human cultured cells. We should like to thank Mr R. Penketh ar.d Mr G. F’razer for their help in the early stages of this project, and St,eveJeremiah and Lorraine Evans for enzyme analysis in the hybrids. We are also most grat,efillto tile following people who allowed us to use their hybrids: Dr E . Solomon, Dr I. Craig, Genetics Laboratory, University of Oxford ; R. Buckland, MRC Clinical and Population Cytogenetics Unit, Edinburgli ; Dr P. Gormley, formerly MKC Clinical and Population Cytogenetics Unit, Edinburgh ; Dr Rosalie Ber, Inst itut.c trf Turnour Biology, Stockholm ; Dr B. Carritk, Department of Ccnetics, University of Glasgow ; Dr T. bfcagcr, Departrnent of Biological Sciences, University of Warwick.
REFERENCES
BOF~IKW, & &I CROSS, . S. (1974).Differential staining of human and mouse chromosomes in iiiter-spccific ccll Ilybrids. Nature, Lond. 251, 77. U K I ~ ~ E A., R , RICHTERICH, It. & AEBI, H. (1964). Die heterogenitat der kreatiri-kinase. Bioclwwische Zeitschrifi 339, 305. CARIIITT, H., GOLDPAARB, P. I?., HOOPER, M. L. B SLACK,C. (1977). Chroniosomal assignment of R Iinrnaii gcnc for argininosuccinate synthetase expression in Chinese hamster and human somatic cell Iiybrids. I Cell Res. 106, 71. i. W., PRILIPO, L., M ~ L T Z EH. R , Y. & KELLER,C. (1977). Iso-enzymes of creatine phospliokiiiaso in
\\.liitu blood cells. Ezperientia 33, 166. CHIJ,E. H. Y., BRIMER,P., JACOBSON, K. B. & MERRIAM, E. V. (1976).Mammalian cell genetics. I. Seleciioti and characterisation of mutations auxotrophic for 1-glutamine or resistant to 8 azagliailine in Cliitiese Jrnmster cells ,ir~vitro. Genetics 62, 359. DAWSON, D. M., EPPENBERGER, H. M. & %PLAN, N. 0. (1965). Creatine kinase: evidence for a tliincric structure. Biochem. biophys. Res. Cornmun. 21, 346. EVANS,J . , STEEL,M. & ARTHUR,E. (1974). A hernaglutination inhibition technique for detection of iinmminoglobulins in supernatants of human lymphoblastoid cell lines. Cell 3, 153. FELLOUS, M., COUILLIN, P., NEAUPORT-SAUTES, C., F R ~ Z AJ., L ,BILLARDON, C. & DAUSPET. J. (1973). Studies of human alloantigens on man-mouse hybrids : possible synteny between HL-A a i d P systwis. A’ur. J . Immunol. 3, 543. FISHER, R. A., POVEY, S., BOBROW, M., SOLOMON, E., BOYD,Y. & CARRITT, B. (1977). Assignment of tlic M A , locus to chromosome 22. Ann. H u m . Genet., Lond. 41, 151. I T A x i z r s , H. & HOPKINSON, D. A. (1976).Handbook of Enzyme Electrophoresis in Huntan Genetics. Axnstcrdain : Sortli-Holland Publishing Co. ox,J., POVEY,S. & HARRIS,13. (1977). The analysis of malignancy by cell fusion. VII. Cytogctiotic analysis of hybrids between malignant and diploid cells and of tuinonrs derived from thcm. J . Cell iSc.ier,ce 24, 217. .JOSEY, E. A . , GOODFELLOW, P. N., KENNETT,It. H. & BODWER, ‘VV. F. (1976). The independent csprwsion (of H LA and ,8,-microglobulin o n human-mouse hybrids. Xomat. Cell Genet. 2, 483. ~ ~ A I E N I I O F E R ,M. C., DELAIN, F., HANZLICKOVA-LEROUX, A., BOG&A. & DREYFUS, J. C. (1970). Isociizynic studies in human embryonic tissues and cell cultures. Protides Riol. Fluids 18, 103. POVEY, S., SLAUGHTER, C. A., WILSON,D. E., GORMLEY,I. P., RUCKTON, K. E., PERRY, 1’. & BOBEOIV, 31. (1976). Evidence for the assignment of the loci AK,, A K , arid A C O N , to chromosome 9 in man. A n n . Htc7n. (:enet., I a a d . 39, 413. ROSILKI,S. B. (1967). An improved procedure for serum creatine phosphokinase determination. J . Lab. C‘lin. M e d . 69, 696. SANTACHIhRA, A. s., NABHOLZ, kf., MIGGIANO, v., DARLINGTON, A. s.& BUD~VER, 3%’. F’. (1970). C h i t ~ t j e analysis with man-mouse somatic cell hybrids. Nature, Lond. 227, 248. SOLOXON, E., BOBROW, M., GOODBELLOW, P. N., BODMER, W. F., SWALLOW, D. M., POVEY, S. & XOEL,B. ( 1976). Human gene mapping using an X/autosome translocation. Xomat. Cell Genet. 2, 125. SOLOMON, E., SWALLOW, D., BURGESS, S. & EVANS, L. (1979). Assignment of the human acid a-glucosidase gene ( a G L U ) to chromosome 17 using somatic cell hybrids. Ann. H u m . Genet., Lond. 42, 273. T s u N G , S. H. (1976). Creatine kinaso isoenzyme patterns in human tissue obtained a t surgery. C h . Chem. 22, 173.
26
SUSAN POVEY AND OTHERS
HEYSINGEN.V., BOBROR,&I., BODJER, R'. F., GARDINER,S. E., POVEY, S. & HOPKINSON, D. A. nssiznment of some human enzyme loci ; mitochondria1 malate dehydrogenase t o 7 , marinos~-phospiiateisomerase and pyruvate kinase to 15 and probably esterase D to 13. Ann. H u m . G e n e t , Lond. 38, 295. WRETOV. E. J. B PFLEIDERER, G. (1975). Quantitation of creatine kinase isoenzymes in human tissues and sera by an iniiniinnlogical nietliod. CI?ii. Chito. Actu 58, 223.
VAX
( 1973). Chromosome
15
Printed in Great Britain
The expression of creatine kinase isozymes in human cultured cells BY SUSAN POVEY," MARY INWOOD," ALISON TANYAR" AND MARTIN BOBROW?
* MRC Human Biochemical Genetics Unit, Galton Laboratory, University College London and Genetics Laboratory, Department of Biochemistry, Oxford INTRODUCTION
Three isozymes of creatine kinase have been identified in human tissues, the MM form characteristic of skeletal muscle, the BB form which is predominant in brain, and the heteromeric MB form which is most active in heart muscle (Burger, Richterich & Aebi, 1964; Dawson, Eppenberger & Kaplan, 1965). Sensitive immunological techniques and quantitative assay after ion exchange chromatography have detected varying proportions of these isozymes in other tissues (Wretou & Pfleiderer, 1975; Tsung, 1976). It has been reported that the BB isozyme occurs in human foetal fibroblasts (Maienhofer et al. 1970) and also that the MM form may occur in lymphocytes and polymorphonuclear leucocytes (Cho et al. 1977). We have investigated the occurrence of creatine kinase in human lymphoid lines, fibroblasts and also rodent cell lines and human-rodent somatic cell hybrids.
MATERIALS AND METHODS
Human tissues, usually brain, heart and skeletal muscle, were available in the course of other investigations and were stored at - 20 "C. Mouse and hamster tissues were stored in the same way. Human fibroblasts were cultured from small skin biopsies using routine methods, and were maintained in Eagle's MEM with 10 yo foetal calf serum. Human lymphoblastoid lines, which were originally kindly supplied by Dr Michael Steel, were maintained in RPMI 1640 with 20 yo foetal calf serum. Rodent cells and human-rodent somatic cell hybrids were supplied to us in the course of other projects by various laboratories (see Table 1) and were transported as frozen cell pellets. The origins of the various rodent. cell lines tested are given in the references to Table 1, except for the hamster line V79 (Chu et a&.1976). Cultured fibroblasts were examined where possible on the day of harvesting. Other cultured cells were usually stored in liquid nitrogen until they could be analysed. Tissues were extracted by homogenization in distilled water ( l : l ,w/v) using a Silverson emulsifier. Washed pellets of lymphoid lines (approx. 5 x lo7 cells) were mixed with a volume of water approximately equal to the cell pellet and subjected to two short bursts of sonication from a probe sonicator, the sample being surrounded by ice. Homogenates and sonicates were then spun at 12 000 g for 20 min and the supernatant used for the investigation of CK. For the cultured fibroblasts a pellet of approximately 4 x lo6 cells was sonicated in the smallest possible volume of water and the whole sample applied to the gel. Rodent cells and human-rodent hybrids were usually extracted as for lymphoid lines. For the investigation of CK in the various human cell lines starch gel electrophoresis was 0003-4800/79/0000-4304 $03.00 @ 1979 University College London
16
SUSAN POVEY AND
OTHERS
Fig. 1. Starch gel showing creatine kinase BB isozymes in a number of different lymplioid lines. Approximately the same amount of sample from each line was applied to the gel. Elcctroplioresis hist,idine-citrate,p H 7.0. Tetrazolium-linked stain.
carried out using 0.005 M histidine/NaOH p H 7.0 as the gel buffer and 0.41 BI citric acid pH 7.0 as bridge buffer (Harris 8: Hopkinson, 1976). The gel was run a t 2 V/cm for 18 h with cooling. On occasion an alternative buffer system, Tris-citrate pH 7.5, was used; bridge 0.094 11 Tris/0.0235 M citrate, gel a 1 in 5 dilution of this; running conditions as before. For the separation of the human and rodent BB isozymes a citrate phosphate buffer system pH 5.8 n-as used (bridge 0.11 M citrate, 0,245 M phosphate, gel a 1 in 40 dilution of this). The gel was run at 4 V/cm for 17 h with cooling. Detection of creatine kinase was by two standard methods, either using creatine and a negative fluorescent stain, or using creatine phosphate and a tetrazolium-linked stain (Harris 8: Hopkinson, 1976). When the tetrazolium stain was used the concentration of creatine phosphate was increased approximately two-fold over that described. I n general, activity in the cultured cells was rather low and the gels were incubated with the stain for 1-2 h a t 37 "C. Human chromosome identification in the hybrids was by G-11 staining followed by Q-banding as described by Bobrow 8: Cross (1974). RESULTS
Human cells Fibroblasts from 20 individuals (10 adults, 9 foetuses and 1 amniotic cell culture) together with 46 lymphoid lines from 42 individuals were screened for the presence of' CK isozymes. The lymphoid lines included those from 12 patients with malignant disease (mostly Burlcitt lymphomas or leukaemias), from 16 patients with a wide range of other disorders (including infectious mononucleosis, cystic fibrosis and X chromosome abnormalities) and from 14 healthy babies and adults.
C’reatine kinase isozyrnes in human cells
17
Fig. 2. Opposite halves of a starch gel stained with and without creatine in t,he negative fluorescent stain. ( a ) Non-specific isozymes; ( b ) shows in addition true creatine kiriase isozynies in heart, brain and the lymphoid line TAY,. Electrophoresis histidine-citrate, pH 7.0.
Thc BB isoxyme was detected satisfactorily by either staining method, and the bands observed were always dependent on the presence of creatine or creatine phosphate in the stain. In all fibroblasts tested the BB isozyme was present and no consistent differences between adults and foetuses were observed. Seventeen of the lymphoid lines from 16 different individuals mere clearly positive for the CK BB isozyme, but there was considerable variation in the amount of activity expressed (Fig. 1 ) . For example, three separate lines T A P 2, TAY 3 and TAU 4 were started on the same day from the same individual, a patient with acute myeloid leukaemia (Evans, Steel & Arthur, 1974). TAY 4 consistently had very high creatine kinase activity, similar to that of brain, whereas in TAY 2 the activity was much less and in TAY3 it was not detected. CK BB was found in 8 out of the 16 lines from healthy individuals (including both adults and babies), in 3 out of 14 lines from patients with malignancy (including TAY 2 and TAY 4) and in 6 out of 16 lines from patients with other diseases. He\ era1 other enzymes including alkaline phosphatase and malic enzyme show considerable line-to-line variation in activity, but neither of these latter enzymes showed a definite correlation with the expression of CK. However, the lines negative for malic enzyme (exclusively lymphomas and lymphatic leukaemias) have so far also been negative for CK BB and will be reported in detail elsewhere. The M B and M M isozymes were generally investigated using the negative fluorescent stain, since the tetrazolium-linked stain has the disadvantage that adenylate kinase (AK)isozymes also appear on the gel. In the fibroblasts the results were clear and no MB or MM isozymes were observed. Much more difficulty was encountered in the assessment of the lymphoid lines. Varying amounts of components in the MM and MB position were seen in the lines, but these 2
H G E 43
SCSAX POVEY AND OTHERS
did not clep,rntl on tlie presence of creatine in the stain (Fig. 3 ) . Similar results were obtained in all th13t.t~elcctroplioretic systems. The bands in the JIB position depend on the presence of both A'I'P ;iad PEP (pliosl)lioenolp?-rurate)in the stain. They may be non-specific phosphatases, but iond in position to nnj- of the well characterized human phosphstases (Harris & Flopkinson. 1 O i f i ) . The isozj-mes soiuetimes seen in the At31 position are reduced in iwtivity by high \ l ) t ' t ~ lcrnti*ifiigation(100000 g for 1 11) of tlie extracts and also do not require PEP for actij-itj I t ~t'euispo-sible that this actirity is attributable to membrane-bound hydrolases, althotigh the n-aj in which they are revealed by this stain is not clear. Bec,iiiw of tlic tlifficulties with the negative stain the gels were also stained using the tetrazoliuiii-linkctl staiti. Xi1 at tempt was madr. to inhibit the AK isozymes using adenosine monophoy)Ii~ite(tip to 1 mgZ/ni1in stain) as suggested by Rosalki (1967). b u t this was only partially succes.f~il. F,tiiit crea tine lciiiase JIM isozyines could have been obscured by AK isozynies (altho1igl~-lK1 i4 bai-clj detectable in most lymplioicl lines), but the JIB isozyme should have i l k . So ti ~ C of C JIB or JZJI isozyne was seen in any lymphoid line tested.
Creatine Einase isozymes in human cells
19
Fig. 4. Part of a starch gel showing creatine kinase isozymes in hybrids made with the mouse line P G 19. The insert line of the gel is not shown. Electrophoresis citrate-phosphate pH 5.8. Tetrazoliumlinked stain.
Rodent cells
A survey of mouse and Syrian and Chinese hamster tissues, using both stains, revealed a distribution of isozymes similarto that of the human. After electrophoresis in citrate phosphate pH 5.8 the human and mouse BB isozymes were clearly distinguishable, the human being more anodal than either hamster or mouse (Fig. 3). I n view of the findings in the human cultured cells, the study of the cultured rodent cells was centred on the BB isozymes. For this the tetrazolium-linked stain was used as it appeared to be slightly more sensitive and certainly gave sharper bands. The mouse line PG 19, originally derived from a melanoma (Jonasson, Povey & Harris, 1977) had a clear band of CK BB in the same position as the brain enzyme (Fig. 3). The other mouse lines tested, lR, 3T3, A9HT and RAG and the hamster lines V79 and WgSh, appeared to lack creatine kinase. Hybrids
A total of 31 primary independent hybrids were tested for CK. Hybrids made between human cells and PG 19 were clearly scorable for the presence or absence of human CK B polypeptide. Positive hybrids had a three-banded pattern consistent with the dimeric nature of CK BB, whereas negative hybrids had the mouse enzyme only (Fig. 4).Hybrids made with other rodent lines either had no CK or the human CK BB enzyme only (Fig. 5). The right-hand column of Table 1 indicates in which crosses human CK BB was seen in at least one clone. Analysis In the analysis of a differentiated function four classes of human-rodent hybrid can be envisaged, i.e. both parental cells expressing the function, neither expressing it, or expression occurring in the human or rodent parent only. We have studied hybrids in all four groups and 2-2
20
SUSAN POVEY AXD OTHERS
Fig. 5 . Starch gel showing creatine kinase isozymes in three different mouse lines and human-rodent hybrids made with them, showing that human CK BB can be expressed independently of the mouse enzyme. Tetrazolium-linked stain. Electrophoresis citrate-phosphate, p H 5.8. (PG 19 has a characteristic two-banded pattern for AK, and may be a heterozygote).
also some in which the status of the human parent with regard to CK BB was unknown (Table 1). From our results so far it appears that the mouse enzyme in PG 19 is not extinguished by fusion with human cells, and is uninfluenced by human chromosomes. Similarly, the lack of CK BB in the other rodent lines was not affected by fusion. The presence of human CK alone in some of the hybrids made with these cells provides good confirmation that these rodent lines are truly not expressing CK BB and the failure to demonstrate it is not due to technical reasons. I n the few hybrids made with human parental lines known not to express CK (the lymphoid lines MOLT, and RAJI) the human CK BB isozyme has not been seen, but this cannot exclude the possibility that reactivation of human CK BB might occur in such a hybrid. Partly because of the negative results, but also because both RAJI and MOLT, are highly aneuploid these hybrids have been excluded from the subsequent analysis. I n an ideal situation one would analyse each cross separately, but insufficient data are available to do this usefully a t present. Table 2 therefore shows a summary of the segregation of human CK BB in all 28 independent hybrids (excluding those made with RAJI and MOLT,). There is some discordance with all chromosomes. This is not unexpected since on a background where the mouse enzyme is not expressed complex factors may determine the expression of the human gene. Even so, it seems reasonable to assume that human CK BB cannot be expressed in
25
Normal fibroblasts X/I 5 translocation fibroblast I / I I translocation fibroblast I /17 translocation fibroblast Normal fibroblast Normal fibroblast Lesch Nyhan fibroblasts MOLT, lymphoid RAJI lymphoid T lymphocytes BRE lymphocytes
Fisher el al. '977 Solomon et al. '976 Povey et al. I976 Povey et al. I976 Solomon et al. I979 F'ellous et al. '973 Carritt et al. '977 Unpublished (Oxford) Dr Rosalie Ber Jones et al. 1976 Sant,achiara et al. 1970 van Heyningen et al. 1975 Lymphocytes
Human parent
Reference or source
4 I I
I Rmouse
0
2 I0
I
8
0
AgHT mouse P G I 9 mouse 3T3 mouse
2
20
-
A 23 hamster PG 19 mouse
0
I
LMTK- mouse
2
8
3
I
8
2
0
RAG mouse
A23 hamster
5
I
I Rmouse Wg3h hamster
7
p G 1 9 mouse
Rodent parent
No
No Yes Yes
No
Yes
N O
Yes
Yes
Yes
Yes
Yes
Human CK BB seen in hybrid
Parentheses indicate that, although CK was not tested on the parental cells, it was assumed t o be present as no example of human fibroblasts not expressing CK BB has been found.
HORP
AgHT/RAJI TP Z412
MOP
PF
P7A/2
MOG
8z/Az3
81 /wg
DURq
DFP
Name of hybrid
No. of indepen- No. of subdent hybrids clones
Table 1. Expression of human CK BB in human-rodent hybrids
22
SUSAN POVEY AND OTHERS
Table 2. Segregation of human creatine kinase BB and chromosomes as assessed by marker enzymes and/or karyotyping in 28 independent primary hybrids CK/chromosomes ,-----Jc------
Chromosome I 2
3*
4 5
6 7 8* 9 10 I1
I2
I3
14
PGM,/ENO,/PGD/PEPC MDHJACP, GPX PGM, HEX, ME,/PGM,/GLO JIGUS GSR AK,/AK,/ACON,
GOT,
8 8 4
9
5 '4 7 5 9
8 I2
ESD NP
5 16
19 20
ADA
16 I7 I8
21
SOD,
X
G6PD
-/-
I4 7
I0
LDH, LDH,/PEPB MPI/PK APRT GALK PEPA GPI
15
+/+
Enzymes
9
I5 I1
12 15
7
+/-
-/+
5
4
2
3
I
2
3
I
4
5 4 5
6
I
2
8
4
3
2
2 2
7 8 7 7
5 5 7 4 6
4
7 7
0
5 5
4
5 4 7 4 6
5
5
6
9
15
5 5
I 0 I
I0 I
I
5 4 I
3
7 7
9 6 '4 I 2 I1 * Enzyme markers used in only a few hybrids - most of the assessments of these two chromosomes (3 and 8)were from karyotype analysis.
the absence of its structural gene. There are two chromosomes (14 and 1 7 ) which are present in every hybrid expressing human CK, and a further five chromosomes which have been recorded as absent in the presence of CK BB on only one occasion (3, 4, 15, 18, 20). Analysis of subclones from five independent positive hybrids was therefore undertaken. Segregation of CK BB occurred in three of the hybrids. Details of the chromosomes in subclones of two of these are shown in Table 3 (the others were not karyotyped). Definite examples of the retention of CK BB in the absence of chromosomes 3, 4, 15, 18 and 20 were found. The segregation of human CK BB with chromosomes 14 and 17 in subclones of all five hybrids is summarized in Table 4. Three further sets of subclones were tested for CK BB, although the original human-hamster hybrids (PF, see Table 1) had not been tested. The segregation of human CK BB and chromosome 14 in those subclones is also shown in Table 4. These hybrids were assumed to have chromosome 17 because they were originally made by selection for thymidine kinase and maintained in HAT medium. DISCUSSION
If a differentiated function is defined as a gene product found very much more in some types of cells in a living organism than others, creatine kinase BB must certainly be regarded as such. The simplest explanation of the striking differences in creatine kinase BB activity in the cultured cells found here is that this is a reflexion of variation of CK BB activity in their various cells of origin. Although it cannot be excluded that some of the lines may have lost
+
-
DTI-5
-
-
-
-
-
6
-
rt
7
+
-
-
-
10
+
i
11
+ - + -
-
+ -
9
8
.
.
-
+ - .
.
-
-
+ -
*
-
+ + + -
+ + ++ + + +
g
ZI
22
+ + +
20
-
X 74+
- - - + + f + -+/?+/-- + - + - - - + + - - + + - * + - + + + + + - - + + + . + . . +
2
16 17 18 19
++/-+/-+
14 1 5
+ - -
f
-
-
-
13
+ + + + + + + + + + + . . + + + + - +
*
-
+
IZ
of cells; t , chromosome seen in 30 cliromosomo uncertain ; + / - , discrepancy (ciixynie/chroniosorne).
+ , Chromosome seen in >
DTI-5 which was not karyotyped.
< 30 76 of cells; L - , cllromosornc seen in occasional cell;
?, identification of
The presence of human chromosomes was assessed primarily hy lraryotyping but checked by enzyme studies in most cases. The exception was
-
+
+
-
5
- - - - - - - +/+ - - + - - - - + - + - - - -
+
-
+ + -
4
+
3
&
+
2
- + + + - + - - +
- -
-
-
-
+
Z
+
18 DURq DUR4R
I2
I0
Z412 Clones of Z412 4
I
CK
Table 3. Xegregation of CK BB and human chromosomes in subclones of two independent hybrids
3
p.3
w
3
s-
cn
3 m
*.
2
Q
w *.
h
s2.
cl
24
SUSAN POVEY AND OTHERS
Table 4.Segregation of CK BB with NPlchromosome 14 ant1 GALKlchromosome 17 in all subclones tested
-I+ Subclones from five NP/ I4 positive hybrids (pooled) GALK/I7 Subclones from three untested hybrids (a) NP/w GA4LK/I 7
(b) (c)
2
16
0
0
7 7
0 0
0 0 0
NP/ I4 GALK/I 7
I
5
I
0
0
SP/I4 G=~LK/I~
I
0
0
I
0
0
7
CK BB activity as a result of long-term culture, it seems possible that the study of CK BB in hybrids may shed some light on the process of differentiation. Snalysis of human-rodent hybrids described here has suggested that the presence of either chromosome 14 or 17 or possibly both, is necessary for the expression of human CK BB, and there is some evidence that CK BB activity is lost if chromosome 14 is lost. Even in thc cross where both parents are positive for CK BB neither 14 nor 1 7 alone is sufficient for the expression of CK BB. There are also two hybrids related t o the positive JIOG3.6A which have enzyme markers for both 14 and 1 7 and yet do not express CK BB. The simplest explanation of these results is that the structural locus CK, is on chromosome 14 and that other human chromosomes are necessary for its expression. So far it has not been possible to identify the specific chroniosomes involved - this may be because of insufficient data, or because different chromosomes are involved in different lines. Another possible explanation is that we are not looking a t the cffect of particular human chromosomes, but of some factor(s) related to the human contribution whilst not absolutely related to the human chromosomes. It may be relevant that the only 31OG hybrid to express human CK BB appears to have virtually a full set of human chroniosomes. Clearly more data are needed. It is a feature of most human-rodent hybrids that the rodent karyotype remains almost complete whereas the human contribution varies. I n the case of CK BB e~aetlgrthe same o c ( ~ r s , the rodent phenotype being constant and the human variable. This suggests either that diffusiblc ‘activators’ or ‘repressors’ do not exist, or that they are species-specific. It would therefore br useful to look a t some intra-specific hybrids, but the value of these would be greatly increased if a variant of CK BB were used, since the different parental contributions could be assessed separately. Unfortunately, no variants of CK BB have yet been reported.
SUMMARY
The BB isozyme of creatine kinase is consistently present in cultured human fibroblasts and shows great variation in activity in long-term lymphoid lines. One mouse line tested, PG 19, had strong activity, but all other rodent lines tested did not express CK BB. Human and mouse CK BB can be expressed independently of each other in human-rodent somatic cell hybrids.
Creatine kinase isozymes in human cells
25
There is some evidence that the structural locus for CK BB may be on chromosome 14, but the involvement of other chromosomes, especially no. 17, cannot be excluded. The MM and MB isozymes of creatine kinase were not seen in any human cultured cells. We should like to thank Mr R. Penketh ar.d Mr G. F’razer for their help in the early stages of this project, and St,eveJeremiah and Lorraine Evans for enzyme analysis in the hybrids. We are also most grat,efillto tile following people who allowed us to use their hybrids: Dr E . Solomon, Dr I. Craig, Genetics Laboratory, University of Oxford ; R. Buckland, MRC Clinical and Population Cytogenetics Unit, Edinburgli ; Dr P. Gormley, formerly MKC Clinical and Population Cytogenetics Unit, Edinburgh ; Dr Rosalie Ber, Inst itut.c trf Turnour Biology, Stockholm ; Dr B. Carritk, Department of Ccnetics, University of Glasgow ; Dr T. bfcagcr, Departrnent of Biological Sciences, University of Warwick.
REFERENCES
BOF~IKW, & &I CROSS, . S. (1974).Differential staining of human and mouse chromosomes in iiiter-spccific ccll Ilybrids. Nature, Lond. 251, 77. U K I ~ ~ E A., R , RICHTERICH, It. & AEBI, H. (1964). Die heterogenitat der kreatiri-kinase. Bioclwwische Zeitschrifi 339, 305. CARIIITT, H., GOLDPAARB, P. I?., HOOPER, M. L. B SLACK,C. (1977). Chroniosomal assignment of R Iinrnaii gcnc for argininosuccinate synthetase expression in Chinese hamster and human somatic cell Iiybrids. I Cell Res. 106, 71. i. W., PRILIPO, L., M ~ L T Z EH. R , Y. & KELLER,C. (1977). Iso-enzymes of creatine phospliokiiiaso in
\\.liitu blood cells. Ezperientia 33, 166. CHIJ,E. H. Y., BRIMER,P., JACOBSON, K. B. & MERRIAM, E. V. (1976).Mammalian cell genetics. I. Seleciioti and characterisation of mutations auxotrophic for 1-glutamine or resistant to 8 azagliailine in Cliitiese Jrnmster cells ,ir~vitro. Genetics 62, 359. DAWSON, D. M., EPPENBERGER, H. M. & %PLAN, N. 0. (1965). Creatine kinase: evidence for a tliincric structure. Biochem. biophys. Res. Cornmun. 21, 346. EVANS,J . , STEEL,M. & ARTHUR,E. (1974). A hernaglutination inhibition technique for detection of iinmminoglobulins in supernatants of human lymphoblastoid cell lines. Cell 3, 153. FELLOUS, M., COUILLIN, P., NEAUPORT-SAUTES, C., F R ~ Z AJ., L ,BILLARDON, C. & DAUSPET. J. (1973). Studies of human alloantigens on man-mouse hybrids : possible synteny between HL-A a i d P systwis. A’ur. J . Immunol. 3, 543. FISHER, R. A., POVEY, S., BOBROW, M., SOLOMON, E., BOYD,Y. & CARRITT, B. (1977). Assignment of tlic M A , locus to chromosome 22. Ann. H u m . Genet., Lond. 41, 151. I T A x i z r s , H. & HOPKINSON, D. A. (1976).Handbook of Enzyme Electrophoresis in Huntan Genetics. Axnstcrdain : Sortli-Holland Publishing Co. ox,J., POVEY,S. & HARRIS,13. (1977). The analysis of malignancy by cell fusion. VII. Cytogctiotic analysis of hybrids between malignant and diploid cells and of tuinonrs derived from thcm. J . Cell iSc.ier,ce 24, 217. .JOSEY, E. A . , GOODFELLOW, P. N., KENNETT,It. H. & BODWER, ‘VV. F. (1976). The independent csprwsion (of H LA and ,8,-microglobulin o n human-mouse hybrids. Xomat. Cell Genet. 2, 483. ~ ~ A I E N I I O F E R ,M. C., DELAIN, F., HANZLICKOVA-LEROUX, A., BOG&A. & DREYFUS, J. C. (1970). Isociizynic studies in human embryonic tissues and cell cultures. Protides Riol. Fluids 18, 103. POVEY, S., SLAUGHTER, C. A., WILSON,D. E., GORMLEY,I. P., RUCKTON, K. E., PERRY, 1’. & BOBEOIV, 31. (1976). Evidence for the assignment of the loci AK,, A K , arid A C O N , to chromosome 9 in man. A n n . Htc7n. (:enet., I a a d . 39, 413. ROSILKI,S. B. (1967). An improved procedure for serum creatine phosphokinase determination. J . Lab. C‘lin. M e d . 69, 696. SANTACHIhRA, A. s., NABHOLZ, kf., MIGGIANO, v., DARLINGTON, A. s.& BUD~VER, 3%’. F’. (1970). C h i t ~ t j e analysis with man-mouse somatic cell hybrids. Nature, Lond. 227, 248. SOLOXON, E., BOBROW, M., GOODBELLOW, P. N., BODMER, W. F., SWALLOW, D. M., POVEY, S. & XOEL,B. ( 1976). Human gene mapping using an X/autosome translocation. Xomat. Cell Genet. 2, 125. SOLOMON, E., SWALLOW, D., BURGESS, S. & EVANS, L. (1979). Assignment of the human acid a-glucosidase gene ( a G L U ) to chromosome 17 using somatic cell hybrids. Ann. H u m . Genet., Lond. 42, 273. T s u N G , S. H. (1976). Creatine kinaso isoenzyme patterns in human tissue obtained a t surgery. C h . Chem. 22, 173.
26
SUSAN POVEY AND OTHERS
HEYSINGEN.V., BOBROR,&I., BODJER, R'. F., GARDINER,S. E., POVEY, S. & HOPKINSON, D. A. nssiznment of some human enzyme loci ; mitochondria1 malate dehydrogenase t o 7 , marinos~-phospiiateisomerase and pyruvate kinase to 15 and probably esterase D to 13. Ann. H u m . G e n e t , Lond. 38, 295. WRETOV. E. J. B PFLEIDERER, G. (1975). Quantitation of creatine kinase isoenzymes in human tissues and sera by an iniiniinnlogical nietliod. CI?ii. Chito. Actu 58, 223.
VAX
( 1973). Chromosome
