BiskDiv. ,Skr. S e v . HavUrzdevs., 14 : 51-60
POLYMORPHISM O F SERUM TRANSFERRIN I N COD
Institute of Marine Research, Bergen I N T R O D U C T I O N
An ironbinding protein in human serum lvas found by HOLMBERG ancl LAURELL (1945). This protein binds ferrous ions in such a way that the iron does not react with ad-dipyridyl, and is now usually called trans- ferrin (HOLMBERG and LAURELL 1947, GIBLETT, HICKMAN and SMITHIES
1959).
Inherited variations in this protein has been revealed by the use of electrophoretic techniques (TISELIUS 1937, SMITEIIES 1955) in numerous mammalian species (SCHMID 1961, COOPER and SHARMAN 1964), in chicken (OGDEN, MORTON, GILMOUR and MCDERMID 1962) and in pigeon (MUELLER 196 1, MUELLER, SMITIIIES and IRWIN 1962). Transferrins have also been found in sera of amphibians and reptiles (DESSAUER and Fox 1964).
The transferrin polyiliorphism in species studied has its origin in a genetic system with two or more codominant autosomal alleles, each controlling one or more bands in the electrophoretic pattern. I n the macaque, n/Pacaca mulatta, ten inolecular forms, each controlled by one gene, ancl 24 phenotypes have been detected (GOODMAN, KULKARNI, POULIK and REICLYS 1965). The transferrin types in Norwegian reindeer are controlled by eight alleles and each allele produces two bands (BRBND 1964).
Fish proteins have been studied by HAMOIR (1955), who states that the main characteristics of fish protein are satisfactoriily defined, ancl that a coinparison of the highest and lowest classes of vertebrates is possible from the point of view of protein composition. The comparative study of the electrophoretic pattern of plasma or serum sho~vs, however, that the lowest vertebrates apparently do not have cominon characters inter se or with the mammalian ones (HAMOIR 1955). SANDERS (1964) examined the electrophoretic patterns in sera of three trout species. I n two species he found characteristic protein fractions for the particular species. The rainbow trout, however, exhibited either 6 or 7 fractions. The amount of protein differed both inter- and intraspecificly, but he suggested that the amount present varied with age, sex, species, and diet. KHAILOV
( 1 9 6 2 ) studied the protein variation in sera from cod, plaice, and I~addock in the autumn. He found distinct differences between the species and individual variations of the fractions with time.
A single transferrin band was demostrated in jack (Cal-anx sexfasciatus) by BLUMBERG ( 1 9 6 0 ) . CREYSSEL, SILBERZAI-IN, RICARD and MANUEL ( 1 9 6 4 ) found polymorphism of transferrins in carp (Cyprinus carpio) by starch gel electrophoresis, and FINE, DRIHLON, AMOUCI-I and BOFFA ( 1 9 6 4 ) detected the presence of several transferrin types in eel (Anguilla anguilla) by paper electrophoresis. However, too few carp and eel individuals were studied to decide the heredity with sufficient certainty.
The present paper describes transferrin patterns in cod (Gadus morhua) and discusses the genetic basis for the observed differences. 'The inves- tigation was carried out to identify individual genetic characters for studies of cod populations. Haemoglobin polymorphism has been demo- strated in cod, using agar electrophoresis of freshly prepared oxyhaemo- globin (SICK 1961).
M A T E R I A L A N D M E T H O D S
T h e samples used in the present study are selected from a material collected in 1964 and 1965 for population studies. Unfortunately, part of this material was destroyed before the sera could be analyzed in the laboratory. The sa~nples used are from panmixed populations and from different geographic areas with different frequency of the characters ttudied from one sample to another. As indicators for panmixing, haemo- globin types (SICK 1961) and otolith types (ROLLEFSEN 1933) were used.
Table 1 gives a survey of the material containing six samples, with sera from 682 individuals, listed in a geographical order, starting with the
Table 1. Date, locality, and gear used, of the collected samples.
Number of fishes in
sample Date of
sampling 18 Dec. 64 4 Dec. 64 1 Dec. 64 26 Oct. 64 1 Nov. 64 19 Nov. 64
Hordaland, coastwards Trap-net 100
Smala, coastwards 100
Helgeland, coastwards -))- 98
Nordskot, Vestfjorden Varangerfjorden
Bear Island N73O55'E18O15' Trawl 148
I
Locality
I
Gear
sample Croin Hordaland m the south and ending with tlie sample colIecteci 27 nautical miles SW of Bcar Islancl l r i the north. Cod of different sex ancl ega are replesented in the material, the majority being 4 to 5 years old.
Blood was obtained by puncture of 6ziZ6us arte~zosus in live cod. After clotting overnight at 2-4OC the samples were centrifugcd and the scra pipettecl off. The material collected near Bergen was exanlined fresh, whereas tlie scra from the northernmost locali~ies were shipped to the laboratory in frozen state (-25°C). Without suffic~ent capacity for freezing, preservation of the samples from clistant fishing grounds is a major problem. I n the present investigation cod sera could not be stored in a refrigerator for more than five clays ~vithout clestruction of the trans- ferrin n~oleculer. However, frozen sera gave reliable results even after t~ o or three months, although fresh sera gave a better result.
The electrophoretic technicluc xsas a combination of Giri's inethocl (1956 a, b), inodifiecl by SICK (1961 ancl personal communication). T h e most effective buffer system to enhance indiviclual differences among the protein types was 30.25 g/l Tris (Tris(hydroxymetl~yl)-amiiiolnethane), 3 g/l E D T A (EthyleneciiaminetetraAcetic acid), and 2.3 g/l boric acid (AARONSON and G R ~ N M ~ A L 195 7). The gel was made by illixi~lg 2
%
starch (BDH) ancl 0.8% agar (<<Ionagar)) No. 2, Oxoid) in the buffer and heating for half an hour at 9G°C in a ~vaterbath xvhile stirring gently.Microscope slides covered by 2 ml 01 gel were kept in a refrigerator for a quarter of an hour before use.
Each electrophoretic run lasted for two liours. Voltages of 65 to 70 V bet~veen the encls of the filterpapers were applied, giving currents of 6 to 7 mA for each slide. Six slides, each with two specimens, xvere placecl in the apparatus in each run. After fixation the proteins were stained wit11 A~niclo Black 10B ancl tlie gel dried.
I n cases where the transferrin patterns have been clifficult to deter- mine, usually poorly preserved samples, the sera \\?ere tested in a seconcl run. Tliis run often gave a goocl result, provided that particular care xvas taken to keep the temperature of the gel low.
I n orclcr to identify the transferrin bands the method of GIBLETT el nl. (1959) ~ i a s usecl with some nroclifications. Approximately 0.1 ml of ferrous citrate of specific activity (Fe
57
100 pC/ml were added to 0.2 1111 selected sera. After half an hour at 2°C the sera were subjected to starch/agar gel electrophoresis.
The numbers of the different pattern? observed in each of the six samples have been compared with numbers expected according to the Hardy-Weinberg law of genotype distributions in large ranclorn mating populations. The differences between observed and expectecl distributionr were tested by applying a chi-square test.
Fie. 1. Starchiagar gel electrophoretograms of cocl sera. T h e arrow inclicates the site of sera application, numbers to the left state the number of fish in the sample and the samp'e numbrr, and letters to the right the type of pattern. A11enlarged schematic drawing of all components is shown a t the bottom of the figure.
R E S U L T S
Fig. 1 shows ten different patterns made up of eight slides from dif- ferent runs. An enlarged schematic drawing of all the components is seen a t the bottom of Fig. 1. I n all cases the proteins move towards the anode.
All patterns have a fast moving coinponent (albumin) which frequently reaches the filter paper. Slower moving components of uncertain nature
are seen as one or two failit bands next to the albumin. The components which are studied in this paper are five distinct bands called A, R, C', C, and D ; A being the fastest iiloving cornponcnt. The hands appear alone or two together, ancl thus fifteen coiiibinations are possible. Of these eleven have been found. The bands are represented in the serum plienotypes: A h , AB, AC, AD, RB, BC, BD, CC, CD, ancl C'C, The band D alone has also been found. The niobility for the five components are clearly different, the C' band, however, being only slightly faster than the C band under the electrophoretic conditions used in this investigation. Thc distance between the stained balicls A artcl B is only slightly longer than the distance between B and C, tvliile this distance is essentially longer than the distance betlveen the bands C ancl D.
The intensity of tlie bands vary, ancl consequently the content of transferrin in cod sera differs from one individual to another. Generally the one band patterns are stronger than thc t ~ v o band patterns. However, visible variations in tlie amount of protein according to season of the year have not been dctected.
T h e present material contains speciinens of different sex and age, and the components A, B, C, and D are represented in every agegroup from 3 to 16, in rnales and females, and in immature as ~vell as mature cod.
Fig. 2 shows four slides with tlie three patterns AC, BC and CD com- pared with autoradiograins of the same slicles. The Fe5" is bound to the bands described above, ancl variable quantities of radioactivity appear also in the albuinin component except in one case. No radioactivity is bound to the faint bands of uncertain nature.
A hypothesis involving five coclominant alleles Tf", TIB, TfC', TfC, ancl TfD may be adopted to explain the transferrin variation observecl.
This hypothesis gives the homozygotes TfATf", Tf"Tf13, TEC'TfC', TfCTfC, and Tf"Tfn which are assumed to be responsible for the phenotypes AA, BE, C'C', CC, and DD, giving the hetrozygotes TfATfU, Tf"TfC', Tf iTfC, TfATf", TfBTf", TfUTfC, Tf13Tf", TfC'TfC, TfC'Tfn, and TfCTfD for the phenotypes AB, AC', AC, AD, BC', BC, BD, C'C, C'D, and CD. The distributions of the different transferrin patterns from six samples are presented in Table 2 together xvitll the expectecl dis- tributions according to the Hardy-Weinberg law. The colunin <(Not rep.)), contains expected totals of the notrepresented types AC', BC', C'C' and
C'D.
-
T h e chi-square value for the observed distributions and the A
-
expected Hardy-Weinberg distributions in samples from Helgeland,
--
Vestfjorden, and Bear Island, gives a probability between 0.7 and 0.5 when d.f. =3. According to tlie test it is therefore no significant deviation between the observed ancl expectecl distributions in these samples. Thc
Fig. 2. Starchlagar gel electrophoretograms of trailsferrin in cod sera (left: localizetl from a ~ ~ t o r a d i o g r a m s (right). Arrows show the site of serum application, alicl letters indicate the type of pattern.
samples from Horclalancl, S i n ~ l a and V a r a n g c r f j ~ ~ d are omitted because of the low number in the classes of the expectecl diqtributions.
The gene T f q o m i n a t e s the material, particulary in samples fro111 southern Norway, and the frequencies of Tf-' (cf) seem to increase northwarcl, ~vhile the frequencies of the Tf" allele fluctuate between 0.1000 and 0.1 735. Tf" a ~ c l TfD are rarc in the areas investigatecl.
D I S C U S S I O N
The present investigation has clernolistratecl that cod transferrins are co~ltrollecl by codomiliallt alleles.
The transferrin patterns consist of one or two strong bancls a.11~1 was not found to vary wit11 the age of the fish. This is in contrast to the variation of the Iiaeinoglobin patterns in salnion (KOCH, B E R G S T R ~ M and EVANS 1964), which was found to be ontogei~etic. T h e amount of the cod transferrins may vary wit11 time, though it has not been possible to detect any visible variation by the method used in this stucly. I<I-~AILOV (1962) found variations in the amounts of diflerent fractions during
Table 2. Distribution ol transferrin phenotypes observed in samples of cod from diflerent areas compared with the expected distribution 01 gcnotypcs according to the Hardy-Weinberg law.
Sample
Transferrin patterns Observed gene frequencies
I
Ah A B A c ADm I
B cI
BD Ij
CC/
CD DDI
CT.02 exp. 0.16 .83 .42
3 2.54 7 6.40 14 12.86 13 12.76 12 0.30 1 1.08 4 3.12 3 2.20
1
0.08 0.16 0.18 0.21 8 7
exp. 3.57
5 0 0.20
2 7.30 29.37/ 2.18
48.59 i.53 0.01 t
63 1
60.34 8.94 0.33 1 6
3.73
7 4 2
71.40 4.23
48
50.00 1.43
I ,052 ,137 ,797 ,013
49 .102 .I21 .765 .012
2 1.00 1 1.82 2 2.95 4 2.88 2
1 . 2 O l 5 . 3 1 27 30.01
0.24 3 2.22
1 3 3
16.90 23 21.60 26 24.29 31 33.49 15
0.50 1 0.54 0.35 0.55
the autumn in sera of cocl, but he used colourometry to evaluate the different fractions.
Tlie samples were selected to ascertain that the material was froni paninixed populations. Sainples .ivith distributions of haemoglobin types not in accordance ~ v i t h Hardy-T~\'eiiiberg law were not used, nor samples with niixed types of otoliths. The material presented was collectecl during the autumn of 1964, and only one of seven samples has been omitted from the report.
The transferrin~ are well separated and the patterns of different runs can easily be compared with each other when using this starchlagar gel electrophoresis. The determination of the pattern is not difficult for fresh sera, and the teclinique iq quick ancl siniple, ancl large samples can be treated.
The deterinination of the phenotypes has been difficult for poorly preserved samples. I1 a satisfactory result was not obtained after a second, or perhaps a third run of a serum, the specimen was rejected. Since all sera in a saniple were preserved in the saine way, they also were either used or discarded together. Methodical selection of the material therefore has been avoided.
I t has not been possible to test the rare C' protein autoradiographi- cally. I n this report, however, the band C' is interpreted as a very rare ferrous-binding protein because of the appearance of the band: both the position ancl the strength are comparable to characteristics of the other transferrins. Together with A, B, or D, this band could be difficult to distinguish from AC, BC, or CD. Since C' is rare, however, this is a minor problem, and a higher frequency of C' would primarily give a higher liuinber of the pattern C'C.
Transferiins have earlier been ideiitifiecl in jack, carp, and eel, respectively, by BI~UMBERG ( 1960), C R E Y ~ S E L et al. (1 964), and FINE et al. (1964). The transferrin patterns studied in this report, however, cannot be cornparecl in detail ~ v i t h those, because different techniques have been used. The position and intensity of a band clepend upon the electrophoretic conclitions usecl. The transfcrrin patterns in carp (CREYSSEL et nl. 1964) and in eel (FINE et al. 1964) seem to a great extent to be like the patterns in cocl, but genetic studies confirming the patterns in carp and eel have apparently not been undertaken. O n tlie other hancl, the inherita~lce of transferrins in cod as a polyallele system without clolninance is in strict accordallce with results from studies of higher vertebrates (BECI~MAN 1962, OGDEN et al. 1962).
The variable ironbincling capasity of the albumin component can not be explainecl by the results in this study. However, the ironbincling cod albumin will be investigatecl.
The transferrin patterns in cod, which in any case can be separated into four different molecular types named TfA, TfB, TfC and TfD (ac- cording to COIIEX AND S~IREFFLER 1961) represent four codoininant alleles Tf', Tf", TfC, ancl Tf". A fifth molecular type called TIC' may represent a fifth allele Tf".
SUI\/IMARY
Sera from 682 cod froin six different localities along the Norwegian coast and in the Barellts Sea have been investigated by starchlagar gel electrophoresis. Eleven different patterns coniposed by five molecular types have been found. By autoradiography it has been demonstrated that four of them are irollbindiilg proteins. T h e distributions of the observed transferrin pattcrns were in agreement xvith distributions expected according to the Hardy-Weinberg formula, implying that the bands have their origin in five codominant alleles.
A C I C N O I V L E D G E M E N T S
For help during the sampling I oxve a dept of gratitude to Mr.
T. NILSEN, and to skippers and crews of the vessels: <<Asterias)) and trHekktind)). I a m also greatly indebted to the workers ancl staff of Levendcfisk-gruppen and Hordaland Fiskesalslag S/L for their CO-
operation.
I wish to thank my technician, Mrs. AA. VALEN, for loyal and clever work, Cancl.rea1. 6. N ~ V D A L for many profitable discussions, and Dr.
K. SICK for Iielpf~~l suggestions regarding methodical problems. I an1 also grateful to Dr. P. ENGER and Dr. M. BRRND for criticism and help by improving my EngIish translation.
R E F E R E N C E S
AARONSON, T. and GRONWAL, A. 1957. Impro\~ecl separation of ser~un proteins in paper electrophoresis. A new electrophoretic buffer. Scmzrl. 3. Clitz.Lnb.I?zuest., 9 : 338- 341.
BECKMAN, L. 1962. Slow ancl fast transferrin variants in the same pedigree. i\i'ntz~?-e, Loizd., 1.94 : 796-797.
BLUIVIBERG, B. S. 1960. Biocllemical polymorphisms in animals : haptoglobins and -_
transferrins. P1-oc. Sac. eqb. Biol. il/led., I04 : 25-28.
B R ~ N D , M. 1964. Genetic studies on serum transferrins in reindeer. Hereditas, 52 : 181- 188.
COHEN, B. L. and SHREFFLER, D. C. 1961. A re\iisecl nomenclature for the illouse trans- ferrin locus. Genet. Res., Canz., 2 : 306--308.
COOPER, D. and SHARMAS, G. B. 1964. Transferrin variation in kangaroos. J ~ ( J ~ Z L T ~ , Lotzd., 203 : 1094.
CREYSSEL, li., SILBERZAHN, P., RICARD, G. and MASUEL, Y . 1964. Etude du strurn de carpe (G~,prilzzls c(lr/)fo) par electrophordsc en gel d'amiclon. Bull. Sac. Clzinz. b i d . , 46 (1) : 149-159.
DESSAUER, K. C. ancl Fox. W. 1964. Electrophoresis in taxonomic studies illustrated by allalyses of bloocl proteins. Pp. 625-647 in Leone, C. A. eel. Taxononzic biochelrzi- st?? and serology. New York, The Ronald Press Con~pany.
GIBLETT, E. R., HICICMAX, C. G. and SMITI-IIES, 0. 1959. Serum transferrins. A'ature,Lond., 183 : 1589-1590.
FINE, J. -M.; DRIHLON, A.; AMOUCH, P. and BOFFA, G. 1964. Existence cle groupes skricjues chez A~zguilla a~zguilla L. Mise en Cvidellce par Clectrophorese et auto- racliographie cle plusieurs types de transferrines. C.r. Izebd. S'iarzc. Acad.Sci., Paris, 258 : 753-756.
GIRI, K. V. 1956 a. Zone electrophoresis of serum proteins in agar gel. jVntur- wissetzschafetz, 43 : 36.
- 1956 b. A simple agar-electrophoresis apparatus ancl technique. Ibid., 43 : 232-233.
GOODMAN, M., I ~ U L K A R N I , A., POULIIC, E. ancl REKLYS, E. 1965. Species ancl geographic differences in the transferrin polymorphism of macacjues. Scielzce, A(. Y., 147 : 884--886.
HAMOIR, G. 1955. Fish proteins. Aud. Protein Ctzeln., 10 : 227-228.
HOLMBERG, C. G. alld LAUIZELL, C. B. 1945. Studies on the capacity of serurn to bind iron. Acta plyisiol. scatzd., 10 : 307-319.
- 194.7. Investigations in serum copper. I. Nature of serum copper and its relation to the iron-binding protein in human serum. Acta chem. scand., I : 944-950.
ICIIAILOV, K. M. 1962. Elektroforeticheskoe issledovanie belkov plazy krovitreskovykh ryb. T r z ~ d j ~ inunnu~zsk. biol. Inst., 4 (8) : 202-207.
ICOCH, H. J. A., BERGSTR~M, E. and EVAXS, J. C. 1964. The microelectrophoretic sep- aration on starch gel of the haemoglobins of Salnzo salnr L. Meded.. IT. vlaam. Acad., 26' (9) : 1-32.
MUELLER, J. 0. 1961. Transferrin variation in the serum of pigeons and doves. Ilninutzo- gelzet. Lett., 2 (5) : 17--18.
MUELLER, J. O., Smithies, 0. ancl Irwin, M. R. 1962. Transferrin variation in Colztm- bidae. Genetics. Prilzcetolz, 47 : 1385-1 392.
OGDEN, A. L., MORTON, J. R., GILMOUR, D. G. and MCDERRIID, E. M. 1962. Inheri- ted variants in the transferrills ancl conalbumins of the chichen. ~Vature, Lolzd., 193 : 1026-1028.
ROLLEFSEN, G. 1933. The otoliths of the cocl. FiskDir. Skr. Ser. I$avUlzders., 4 (3) : 1-14.
SANDERS, B. G. 1964. Electrophoretic studies of serum proteins of three trout species and the resulting hybricls within the family Sabnolzirlne. Pp. 673-679 in Leone, C. A. ed. Taxolzolnic biochelnistry atzrl serology. New York, The Ronald Press Company.
SCHMID, D. O., 1961. Uber transferrine bei Tieren. Eur. Soc. Haernat. congr. No. 8, Wielz, Proc. 1 (1 75) : 1-6.
SICK, I<. 1961. H~nloglobin polymorphism in fishes. .Nature, Lolzd., 192 : 894-896.
S ~ r ~ ~ ~ r e s , 0. 1955. Zone electrophoresis in starch gels : Group variations in the serum proteins of normal human adults. Bioch. J., GI : 629-641.
T~sar.rus, A. 1937. A new apparatus for electrophoretic analysis of colIoicla1 mixtures.
T~rozs. Fflrn(fgji ,Ebc., 33 : 524-53 1.
Received 29 Novenlber 1965 PI intecl 15 December 1966
