Ann. Hum. &net. (1981), 45, 13S-141
`Printed in &eat Britain
`
`135
`
`Chromosomal localization of a single copy gene by
`in situ hybridization - human p globin genes on the
`short arm of chromosome 11
`
`BY S. MALCOLM, P. BARTON, C. MURPHY AND M. A. FERGUSON-SMITH
`Biochemistry Department, Queen Elizabeth College, London, W8,
`and Institute of Medical Genetics, Yorkhill, Glasgow
`
`SUMMARY
`1. The localization of the /3 globin genes by in situ hybridization to fixed chromosomes is
`described.
`2. The probe used was a [3H]cRNA copy of a genomic clone containing in total 4 4 kb of DNA
`and including the p globin gene.
`3. The evidence for the localization of the gene comes from three pieces of data, (a)
`Chromosome 11 is labelled to double the extent expected if the grains were randomly
`distributed, (b) the extra grains above background are clustered on the short arm of 11 close
`to the centromere, and (c) the absolute number of grains observed is very close to that predicted
`for a probe of that length by comparison with ribosomal genes. The localization is in agreement
`with that obtained by other methods.
`4. This method could be extended to any gene for which a genomic clone containing a t least
`5 kb of single copy DNA is available.
`
`INTRODUCTION
`In situ hybridization to fixed metaphase preparations from cultured lymphocytes has been
`used to locate several repetitive gene families in man such as 28s and 18s ribosomal RNA genes
`(Evans, Buckland & Pardue, 1974), 5s genes (Steffensen et al. 1975) and histone genes (Chandler
`et al. 1979). This method of gene localization can be used directly on normal chromosome
`preparations and avoids the problems associated with somatic cell hybrids which are time-
`consuming to prepare a d may contain undetected chromosomal rearrangements or fragments.
`The use of chromosomal translocations allows very precise localizations to be made (Fennel1
`et al. 1979).
`The method has previously been restricted to repetitive genes because of the extremely small
`quantity of DNA found in a single chromosome and the maximum possible specific radioactivity
`of the RNA or DNA used as a gene probe in hybridization. However, the availability of cloned
`genomic fragments containing structural genes and the intervening and surrounding sequences
`(Lawn et al. 1978) means that the length of DNA sequence hybridized for single copy genes may
`approach that for repetitive families. We have used an RNA transcript from a restriction
`fragment from one of the best characterized systems, the p globin gene, to confirm both the
`localization of this gene on chromosome 11 and the regional localization close to the centrornere
`on the short arm.
`
`0003-4800/81 /oooO-4516 $01 .OO 0 1981 University College London
`
`GENENTECH 2017
`GENZYME V. GENENTECH
`IPR2016-00383
`
`

`
`136
`
`S. MALCOLM AND OTHERS
`
`METHODS
`
`Chromosome preparations
`Metaphase chromosomes were prepared by culturing normal lymphocytes with phytohaem-
`agglutinin as previously described (Ferguson-Smith, 1974). Before hybridization chromosomes
`were treated with a 1 % aqueous solution of lipsol detergent and stained with Leischman's stain
`for photography.
`
`I n situ hybridization
`After treatment with RNAase A (Sigma) at 100 pg/ml at 37 O C for 30 min, chromosomes were
`denatured in 60% formamide, 5 mM HEPES, 1 mM EDTA, pH 7-0, at 55 OC for 15 min. Ten
`ng of [3H]cRNA (specific activity 1.8 x lo* dpmlpg) in 5 pl of hybridization buffer (50 yo
`formamide, 0-6M-NaC1, 5mM HEPES pH 7.0) were incubated with each slide a t 43 "C
`overnight. Slides were then treated with RNAase A at 10 pg/ml, 37 OC, 30 min to remove
`non-hybridized or mismatched labelled cRNA and washed thoroughly in 2 x SSC. After
`dehydration through a series of alcohols, the slides were dipped in Ilford K 2 Nuclear Emulsion
`and exposed a t 4 O C for 40 days. The slides were then developed and the cells that had previously
`been photographed were re-examined with a microscope for silver grains.
`Complementary RNA. [3H]cRNA was made by the method of Jones (1974). ATP, CTP and
`UTP were radioactively labelled with [3H] and the specific activity of the cRNA was
`1-7 x lo8 dpm/pg.
`
`RESULTS
`
`Hybridization probe
`The recombinant plasmid (HPIS) used to prepare the cRNA gene probe contains the human
`/3 globin gene and its intervening and surrounding DNA sequences to a total length of 4.4
`kilobases (kb). This plasmid, kindly provided by Dr Tom Maniatis of California Institute of
`Technology, was prepared by inserting a Pst I restriction fragment containing the /? globin genes
`into plasmid PBR 322. A full restriction map of this region of the chromosome is shown in
`Bernards et al. (1979). This fragment hybridizes to a single band of restricted DNA in Southern
`blot hybridization at a final wash of 0.1 x SSC, 65 "C (T. Maniatis, personal communication, and
`our unpublished data).
`
`Chromosomal distribution of grains
`Human metaphase cells were banded and stained with Lipsol/Leischman's and photographed.
`The position of each cell was noted and, after in situ hybridization and autoradiography, the
`same cells were relocated by microscope and scored for autoradiographic silver grains. This
`allows the unequivocal identification of each chromosome. Fifty-five cells were karyotyped and
`the grains falling on each chromosome determined (Table 1). Since the cells have been selected
`for the quality of the metaphase preparation prior to in situ hybridization, observer bias in cell
`selection cannot occur. An allowance for the different lengths of the chromosomes has been made
`by dividing the number of grains occurring over each chromosome by the percentage of the total
`genomic DNA found in that chromosome (Ferguson-Smith, 1974). This gives the number of
`grains per unit length for each chromosome where a unit length is 1 O/" of the DNA of the human
`
`

`
`Chromosomal localization of a single copy gene by in situ hybridization
`
`137
`
`Table 1. Total grain counts per chromosome in 55 cells analysed
`
`Chromosome
`I
`2
`3
`4
`5
`6
`7
`8
`9
`I 0
`I 1
`I 2
`'3
`14
`'5
`16
`I7
`I 8
`19
`20
`21
`22
`
`x
`Y
`
`Total no.
`of grains
`72
`61
`63
`54
`58
`5'
`42
`43
`5 1
`47
`93
`53
`34
`36
`37
`49
`31
`38
`26
`21
`I 2
`22
`23
`8
`
`Yo of genome
`8'47
`7.76
`6.56
`613
`5'58
`5'65
`5'00
`4'77
`473
`435
`435
`416
`3'59
`3.28
`3.1 I
`3.1 I
`3.02
`2'73
`2.58
`2.3 I
`1.90
`1.69
`5.14
`2.09
`
`Grains/unit
`length
`8.50
`7.86
`9.60
`8.8 I
`'039
`9'03
`8.40
`9.0 I
`1078
`10.80
`21.38
`12.74
`947
`10.97
`I 1-89
`15.75
`1026
`13'92
`I O . 0 8
`9'09
`631
`I 3.0 I
`8.94
`7'65
`z = 10.61
`u = 3.06
`
`Grains per
`unit length
`
`4 ;;I
`
`2
`
`1
`
`2
`
` -
`
`3
`
`4
`
`5
`
`
`
`I1
`12
`Chromosome
`Fig. 1. Distribution of grains per unit chromosome length (1 Yo of the genome). The dotted line indicates
`the grains expected if grains are randomly distributed.
`
`.-- 1
`
`13
`
`14
`
`I9 20 21
`
`22
`
`X
`
`Y
`
`

`
`138
`
`S. MALCOLM AND OTHERS
`
`Number
`of grains .
`18-
`
`16-
`
`I2-
`
`10- ':I
`14- r
`
`2
`
`L
`
`P
`
`
`
`Fig. 2. Distribution of grains along chromosome 1 1 . The arrow indicates the position of the centromere.
`p and q represent short and long arms respectively. The chromosome was divided into ten equal-sized
`segments.
`
`diploid genome (Fig. 1). A calculation of the grains/unit length expected if all the observed
`grains were due to random background has been carried out by dividing the total number of
`grains by the total number of length units. This value is shown as a dotted line in Fig. 1.
`Chromosome 11 shows twice as much hybridization per unit length as would be expected if
`all the grains were distributed randomly (see Fig. 1) and no other chromosome shows anything
`approaching this level of hybridization. When the grains/unit length for all the chromosomes
`are analysed the value for chromosome 11 falls outside 3-5 standard deviations from the average
`(see Table l), making it extremely unlikely that this is a chance value. No other chromosome
`falls significantly away from the mean although chromosome 16 shows hybridization somewhat
`above background and is the next most heavily labelled chromosome. As expected, the smaller
`chromosomes, which represent a smaller proportion of the total counts, differ more from the
`average than the larger chromosomes.
`
`Distribution of silver grains along chromosome 11
`The distribution of silver grains along the length of chromosome 11 was analysed and is shown
`in Fig. 2. The chromosome was arbitrarily divided into ten equal units and these units merely
`represent the resolution with which we are able to assign grains and do not represent cytogenetic
`bands. Fig. 2 shows that there is a marked concentration of grains adjacent to the centromerc
`in the short (p) arm of the chromosome. All other parts of chromosome 11 show hybridization
`close to the background level expected for random distribution.
`
`

`
`Chromosomal localization of a single copy gene by in situ hybridization
`
`139
`
`Eficiency of hybridization
`It is possible to make an estimate of the number of grains that would be expected after 40
`days exposure for a locus of 4.4 kb present as a single copy and labelled to the specific activity
`used by comparison with known grain counts for the 185 and 285 ribosomal RNA genes in man.
`The ribosomal genes are distributed amongst five chromosomes in man, and as there are
`approximately 50 copies of the ribosomal genes per haploid genome (Young, Hell & Birnie, 1976)
`on average there will be ten copies per chromosomal nucleolar organizer locus, or 70 kb DNA
`sequence. In control experiments we find 1-2 grains per nucleolar organizer after 1 week’s
`exposure for this specific activity cRNA probe from a ribosomal gene recombinant. For the
`globin gene sequence 44 kb in length, we would expect approximately 0 5 grains per chromosomal
`locus after 40 days exposure. Since a total of 55 cells (1 10 chromosome 1 1 s) were scored, a total
`of 55 grains above background would be expected by comparison with the ribosomal gene locus.
`Using the background calculated from all other chromosomes, there is an excess of 47 grains
`over chromosome 11 in the total number of 55 cells, in close agreement with the predicted value.
`
`DISCUSSION
`Human ribosomal genes may be readily detected using short exposure times by in situ
`hybridization and this has proved useful for gene localization and the study of variants (Salmasi
`et al. 1980; Elliott et al. 1980). As an NOR contains on average 70 kb of ribosomal DNA and
`the average mRNA only 1 kb or less it seemed that this method would not be generally
`applicable to the localization of single copy structural genes. It could be calculated that about
`5 kb of hybridizing sequence would give on average 0 5 grains/chromosome under the conditions
`of overall efficiency achieved in in situ hybridization. This is clearly not enough grains to be
`obvious on visual inspection of one cell but should be quite clear if data are pooled from a large
`number of cells.
`The advent of recombinant clones derived from the chromosome rather than from mRNA
`enables fragments of DNA, around 5 kb long, specific to the structural gene but containing
`intervening and adjacent sequences, to be used as a hybridization probe and therefore allows
`structural genes also to be detected by this method. We have used such a fragment, 4.4 kb long,
`to localize the /3 globia gene to the short arm of chromosome 11 close to the centromere, in
`agreement with results of chromosome sorting experiments (Lebo et al. 1979) and somatic cell
`hybrids (Jeffreys, Craig & Francke, 1979; Scott, Phillips & Migeon, 1979).
`Although nick-translated plasmid DNA may be used directly for hybridization and will in
`fact result in an enhancement of signal due to the formation of cross-linked molecules (Malcolm
`et al. 1977), we decided to use a complementary RNA transcribed from the plasmid DNA
`template by E. Coli DNA-dependent RNA polymerase because the higher specific activity and
`the lower background produced by mild RNAase treatment make this more suitable for
`detecting very low levels of grains. We have found in experiments involving ribosomal and highly
`repeated sequences that no differences are found when native or heat-denatured DNA is used
`as the template and therefore there is no specificity of transcription. This is probably because
`the E . Coli RNA polymerase used has only a partial content of u sub-unit. We have also shown
`directly that the whole length of the inserted sequence is transcribed under these conditions
`by hybridization of aS2P-labelled cRNA to filters containing restricted plasmid DNA (data not
`
`

`
`140
`
`S. MALCOLM AND OTHERS
`presented). Chromosome morphology is considerably damaged during the denaturation and
`hybridization procedures and so it is essential for accurate identification that the chromosomes
`should be banded and photographed before in situ hybridization.
`The significance of the slight excess of grains found on chromosome 16 is not understood.
`Although the number of grains involved is very small, it is most striking that the grains are
`clustered in the centre of the short arm. The a cluster of globin genes are located on this
`chromosome (Deisseroth et al. 1977) but their exact position is unknown and it is possible that
`this method is picking up related sequences found near both globin clusters.
`As this method of detection relies on hybridization of sequences surrounding the globin gene
`as well as the structural gene itself, it is important that these sequences should be conserved
`between individuals. Although the agreement of restriction maps of the j3 globin gene cluster
`produced in many different laboratories throughout the world shows that there are very few,
`if any, major rearrangements in normal individuals, Jeffreys (1979) has shown restriction
`enzyme site polymorphisms resulting from a single base change. These seem to be particularly
`common in non-translated intervening sequences. These single base changes will not affect
`hybridization.
`This method should be extendable to any sequence for which a single copy probe containing
`at least 5 kb of DNA is available. It is not even necessary for the probe t o contain the gene
`sequence itself and, in fact, for multiple gene families such as actin, this may prove a great
`advantage. As highly repeated sequences are found close to structural genes Southern blot
`analysis and blotting using total DNA as a hybridization probe (Fritsch, Lawn & Maniatis, 1980)
`would be necessary to show that any particular fragment was suitable for in situ hybridization.
`Previous attempts to locate structural genes by in situ hybridization (Price, Conover &
`Hirschhorn, 1972) used either low specific activity mRNA or complementary DNA copies and,
`as pointed out here and elsewhere (Bishop & Jones, 1972), the length of the probes used and
`their specific activity made the experiments theoretically impossible. Clearly any gene localization
`using in situ hybridization must show, as in this paper, the overall efficiency of hybridization
`and detection. A possible source of confusion in the early experiments was that total mRNA
`was used and this may have contained some repetitive sequences. The use of well-characterized
`cloned probes removes any possibilities of such contamination.
`
`This work was supported by the Medical Research Council. We are most grateful to Dr Tom Maniatis for
`making HPIS available and to Dr E. Boyd for assistance with karyotyping.
`
`REFERENCES
`BERNARDS, R., LITTLE, P. F. R., ANNISON, G., WILLIAMSON, R. t FLAVELL, R. A. (1979). Structure of the
`human Gy-Ay-S-P-globin gene lous. Proc. Natn. A c d . Sei. U.S.A. 76, 4827.
`BISHOP, J . 0. t JONES, K. W. (1972). Chromosomal localization ufhuman haemoglobin structural genes. Nature,
`Lond. 240, 149.
`CHANDLER, M. E., KEDES, L. H., COHN, R. H. t YUNIS, J. Y. (1979). Genes coding for histone proteins in man
`are located on the distal end of the long arm of chromosome 7. Science, N. Y . 205, 908.
`DEISSEROTH, A., NIENHUIS, A,, TURNER, P., VELEZ, R., ANDERSON, W. F., RUDDLE, F., LAWRENCE, J.,
`CREAGAN, R. t KUCHERLAPATI, R. (1977). Localisation of the human a-globin structural gene to chromosome
`16 in somatic cell hybrids by molecular hybridisation assay. Celt? 12, 205.
`ELLIOTT, T., BOYD, E., MALCOLM, S., AITKEN, D. A. t FEROUSON-SMITH, M. A. (1980). A familial variant of
`human chromosome 9 bearing a nucleolus organiser region. Heredity 45, 156.
`EVANS, H. J., BUCKLAND, R. A. t PARDUE, M. L. (1974). Location of the genes codingfor 18s and 28s ribosomal
`RNA in the human genome. Chromosoma 48, 405.
`
`

`
`141
`
`Chromosomal localization of a single copy gene by in situ hybridization
`M. A. (1979). Use of chromosomal
`FENNELL, S. J . , MALCOLM, S., WILLIAMSON, R. & FERGUSON-SMITH,
`translocations with in situ DNA hybridisation to confirm localisation of human 5s ribosomal RNA genes.
`.I. Mad. Genet. 16, 246.
`FERGUSON-SMITH, M. A. (1974). Techniques of human chromosome analysis. La Ricerca Clin. Lab. 4, 297.
`FRITSCH, E. F., LAWN, R. M. & MANIATIS, T. (1980). Molecular cloning and characterisation of the human P
`like globin gene cluster. Cell 19, 959.
`A. J. (1979). DNA sequence variants in the Gy *Y 6 and P-globin genes of man. Ce2l 18, 1.
`JEVFREYS,
`A. J., CRAIG, I. W. & FRANKE, U. (1979). Localisation of the GyAy S and P-globin genes on the short
`JEFFREYS,
`arm of chromosome 11. Nature, Lond. 281, 606.
`JONES, K. W. (1974). The method of in situ hybridisation. In New Techniques in Biophysics and Cell Biology
`(ed. R. H. Pain and B. J. Smith), p. 29. New York: John Wiley.
`LAWN, R. M.,FRITscH,E. F.,PARKER,R. C.,BLAKE,G.&MANIATIS,T.
`(1978).Theisolationandcharacterisation
`of linked 6 and P globin genes from a cloned library of human DNA. Cell 15, 1157.
`LEBO, R. V., CARRANO, A. V., BURKHAN-SCHULTZ, K., DOZY, A. M., Yu, L. C. & KAN, Y. W. (1979). Assignment
`of human /3, y and 6 globin genes to the short arm of chromosome 11 by chromosome sorting and DNA
`restriction analysis. Proc. Natn. Acad. Sci. U.S.A. 76, 5804.
`MALCOLM, S., WILLIAMSON, R., BOYD, E. & FERGUSON-SMITH, M. A. (1977). A comparison of in, situ
`hybridisation techniques for gene localisation. Cytogenet. Cell Genet. 19, 256.
`PRICE, P. M., CONOVER, J . H. & HIRSCHHORN, K. (1972). Chromosomal localisation of human haemoglobin
`structural genes. Nature, Lond. 237, 340.
`SALMASI, A. M., MALCOLM, S., ELLIOTT, T., WILLIAMSON, R., & FERGUSON-SMITH, M. A. (1980). Nucleolus-
`organiser regions in familial extra metacentric human chromosomes. Heredity 44, 63.
`SCOTT, A. F., PHILLIPS, J. A,, & MIGEON, B. R. (1979). DNA restriction endonuclease analysis for localisation
`of human P and S globin genes on chromosome 11. Proc. Natn. Acad. Sci. U.S.A. 76, 4563.
`STEFFENSEN, D. M., PRENSKY, W., MTJTTON, D. & HAMERTON, J. L. (1975). Mapping the human 55 RNA genes
`on chromosome 1 using translocations. Cytogenet. Cell Genet. 14, 434.
`YOUNG, B. D., HELL, A. & BIRNIE, G. D. (1976). A new estimate of human ribosomal gene number. Biochim.
`Biophys. Acta 454, 539.