Journal of Cell Science 109, 773-776 (1996)
Printed in Great Britain © The Company of Biologists Limited 1996
JCS3369
773
Different distributions of homologous chromosomes in adult human Sertoli
cells and in lymphocytes signify nuclear differentiation
Ann C. Chandley1,*, R. M. Speed1 and A. R. Leitch2
1MRC Human Genetics Unit, Western General Hospital NHS Trust, Edinburgh EH4 2XU,
2School of Biological Sciences, Queen Mary and Westfield College, London E1 4NS, UK
Scotland, UK
*Author for correspondence
SUMMARY
Using whole chromosome painting probes for human chromosomes 3, 7, 8, 13, 17 and 21 and X and the probe pHY2.1
for the Y chromosome coupled with fluorescent in situ
hybridization (FISH) analysis, the distribution of chromosomes is reported in nuclei of Sertoli cells of the adult testis
and in stimulated blood lymphocytes. The distribution of
chromosomes in the two cell types is significantly different.
A strong tendency for each pair of homologues to pair is
inferred from the observation of only a single detectable
signal in the majority of Sertoli cell nuclei. The sex chro-
mosomes, by contrast, give two clearly separated signals.
Interphase nuclei in dividing blood lymphocytes, analysed
as controls, also show mainly two separated signals for all
non-acrocentric autosomal pairs, but acrocentric pairs no.
13 and 21 show some tendency to associate, probably
reflecting satellite association.
INTRODUCTION
Our results show significant differences in the distributions
of chromosomes in these two cell types. There is a strong
tendency for somatic association (pairing) amongst all
autosomal pairs analysed in the Sertoli cells of adult men. Our
observations lend support to the idea (Leitch et al., 1994), that
somatic pairing is a feature of non-dividing cell types, and if
this were a prerequisite for normal gene expression, it could
have implications for cells, not only in the testis, but also in
many other human tissues.
An association of homologous chromosomes in mitotic cells
of mammals is not normally observed, although in Drosophila
and other dipteran insects it is a well-established phenomenon
(see Tartof and Henikoff, 1991, for review). Nevertheless, rare
reports of somatic association at interphase, between human
centromeres, have been made, using in situ hybridisation, for
chromosomes 1 and 17 in brain cells (Arnoldus et al., 1989,
1991), and for chromosome 15 in normal and neoplastic cells
of haemopoitic and lymphoid origin (Lewis et al., 1993).
Borden and Manuelidis (1988) showed in nuclei of human
cortical neurons that there is a characteristic repositioning of
chromosomes at interphase associated with cell activity and the
neurological disorder epilepsy. Thus nuclear differentiation is
likely to be of major importance in the activity of the nucleus.
Leitch et al. (1994), have postulated that an association of
homologous chromosomes might be a feature of non-dividing
cells, repositioning arising out of the basic pattern found in
dividing cells. Such ‘nuclear differentiation’, they believe, might
be part of the process of ‘cell differentiation’, somatic pairing
being required for normal gene expression in differentiated cells.
To test this idea, we have analysed the fluorescent in situ hybridisation (FISH) signals produced by chromosome painting of six
selected pairs of human autosomes (nos 3, 7, 8, 13, 17 and 21),
and the sex chromosomes, in human Sertoli cells. These somatic
supporting cells of the germinal epithelium of all mammals proliferate prior to puberty but are non-dividing in the adult testis.
In addition, we have analysed the distribution of the six pairs of
autosomes in dividing blood lymphocytes.
Key words: Somatic association, Sertoli cell, Gene expression,
Lymphocyte
MATERIALS AND METHODS
Chromosome analysis
Autosomal pairs 3, 7, 8, 13, 17 and 21 were chosen for analysis as
these covered the entire size range within the human genome. The X
and Y chromosomes were also investigated.
Tissue cell suspensions
Adult Sertoli cells were found on standard air-dried preparations made
from routine testicular biopsies (Chandley et al., 1994). These were
obtained from individuals C40 and C45, aged 87 and 73 years respectively, who underwent bilateral orchidectomy in the management of
prostatic cancer, from WSM 327, an azoospermic male with physical
obstruction of the vas deferens, and WSM 328 a male with a mean
sperm count of 36.8×106 per ml., i.e. within the range for normal
fertility. The air-dried preparations from each of these individuals
showed abundant germ cells and Sertoli cells, indicative of good spermatogenic activity. Male blood lymphocyte interphase nuclei were
analysed in standard stimulated cultures prepared from two individuals. (Although initial attempts were made to analyse signal distributions in dividing Sertoli cells from foetal testes, these being seen as the
774
A. C. Chandley, R. M. Speed and A. R. Leitch
ideal control for non-dividing adult Sertoli cells, poor quality tissue
preservation in the human abortuses available to us precluded
favourable analysis.) The spreading technique reduced the nuclei to
two dimensions which causes some loss of information and introduces
artefacts, particularly the overlapping of chromatin. However, the
method has the advantage of increasing the speed of data collection
and enabling unequivocal identification of cell type. After spreading,
the nucleus remains as an integral structure and chromosomes do not
separate, as can occur at metaphase. Leitch et al. (1991) examined the
distributions of the two genomes in Hordeum vulgare (Barley) ×
Secale africanum (wild rye) hybrid meristematic nuclei using genomic
in situ hybridization and showed that spread interphase nuclei can be
interpreted for nuclear organization and give comparable data to those
obtained by three-dimensional reconstructions. Thus the data obtained
here enable comparisons to be made between cell types and the overall
trends that occur in many cells, but do not allow the precise position
of chromosomes from any individual cell to be determined.
In situ hybridisation
Whole chromosome painting probes were used to label the autosomes
and X chromosome (Kofman-Alfaro et al., 1994). These were obtained
from Cambio in the biotin and FITC labelled forms. They were
denatured for 8 minutes at 70°C, and then preannealed for 15 minutes
at 37°C. To provide a signal for the Y chromosome, the FITC labelled
long arm heterochromatin pHY2.1 probe (Cooke, 1976) was denatured
for 5 minutes at 70°C and then combined with the paint probes just
before application to the slide. The method of detection depended on
the number of probes applied to one slide. When using three probes
(X,Y and autosome) and the Cambio dual-colour painting kit, the X
was labelled with biotin and detected with avidin Texas Red (red fluorescence), the Y was labelled with digoxigenin and detected with
antidigoxigenin FITC (yellow-green fluorescence) and the autosome
was double labelled with biotin and digoxigenin and after detection
fluoresced orange. When only the X and an autosome pair were studied
the X was again detected with Texas Red and the autosome with FITC.
Signal detection and analysis
Adult Sertoli cells were identified prior to signal analysis, using DAPI
images, and their position noted. The nuclei are large (mean diameter
24.1 µm), ovoid or round in shape, and contain a nucleolus flanked
by two or three chromatin masses (Speed et al., 1993). Blood lymphocyte interphase nuclei appeared large (stimulated) or small
(unstimulated). Analysis was confined to the former which had a mean
diameter of 21.5 µm. All scoring was carried out by confocal
microscopy. Chromosomes occurred in domains, and appeared as
areas of yellow/green, orange or red fluorescence. Sizes of signals
varied according to the size of each individual chromosome pair.
Three types of signal distribution were recognised:
(i) two clearly separated signals (>1 signal diameter apart),
recorded as ‘separate’;
(ii) two signals in close proximity (<1 signal diameter apart),
recorded as ‘close’;
(iii) one signal, recorded as ‘together’.
Fig. 1A-C illustrates the three types of signal distribution observed
in stimulated lymphocytes, using the chromosome 13 paint probe. To
exclude the possibility that single signals might represent monosomy,
checks were made on selected autosome pairs using the PRINS
reaction (PRimed In Situ DNA synthesis) to detect centromeric
alphoid repeats, in combination with the FISH painting.
nuclei for the autosomal pairs, with no evidence for monosomy
(data not given). The findings indicated a strong tendency
towards autosomal chromosome association in adult Sertoli
cells and across the whole range of chromosome sizes. Associated signals for pairs no 7, 13 and 21 are illustrated in Fig.
2A,B and C, respectively. The unassociated X and Y chromosomes are shown in Fig. 2D.
Table 2 gives data for interphase nuclei in dividing blood
lymphocytes. Overall, a mean frequency of 52.7% of nuclei
had homologous chromosomes associated (35% intimately
associated). A close inspection of the data reveals that this
figure is inflated by the contribution made by acrocentric chromosome pairs no. 13 and 21. These showed 74% and 78% of
nuclei with signals associated, 50% of nuclei with intimately
associated signals. Excluding these chromosomes, an overall
mean of 41% of nuclei had associated signals (27.5% of nuclei
with intimately associated signals). These findings indicated no
apparent tendency towards association of homologous chromosome pairs 3, 7, 8, 17 but a tendency towards association
for the acrocentric chromosomes 13 and 21.
A ‘G test’ was used to evaluate the null hypothesis that the
proportion of separate homologous chromosomes (numbers 3,
7, 8, 13, 17 and 21) in each nucleus type (Sertoli cells and
blood lymphocytes) was the same. The distribution of homologous chromosomes is different at a high level of significance
(G=33.87; P<0.001). There was no significant indication of
any effects caused by individual chromosomes (G=3.71; not
significant).
Analysis of the positions occupied by the sex chromosomes
in the 30 nuclei examined showed that, as illustrated in Fig.
2D, the X was almost always peripheral in location and the Y
was almost always central by the nucleolus (data not given).
An incidental finding made when multi-colour FISH was
applied to blood lymphocyte nuclei was that certain chromosomes tended to be seen associated in the interphase nucleus
with certain other non-homologous chromosomes. This was
particularly true of chromosome no. 8 which tended to show a
signal coalescing with the X chromosome signal in more than
70% of nuclei, while by comparison, the no. 17 chromosome
only showed coalescence in 16% of nuclei. The strong
tendency for juxtapositioning of the X chromosome and no. 8
chromosome was also noted for Sertoli cells and pachytene
nuclei of the primary spermatocytes. Although these observations at the present time remain cursory, and will be expanded
in future investigations, they do suggest some specific positioning of chromosomes in relation to each other within the
interphase nucleus which can be tested against models for
chromosome disposition (e.g. see Bennett, 1984).
RESULTS
Table 1 gives the frequency of nuclei in adult Sertoli cells
showing separate, close and together signals, respectively, for
each of the six autosome pairs and sex chromosomes. While
few of the cells showed association of the sex chromosomes
(24.3%), associated signals characterized between 70-83% of
Fig. 1. Human blood lymphocyte interphase nuclei showing the
chromosome no. 13 pair painted. (A) ‘separate’ signals, (B) ‘close’
signals, (C) ‘together’ signals.
Somatic association in Sertoli cells
Table 1. Signal disposition in adult Sertoli cells
Associated (%)
Chromosome no.
3
7
8
13
17
21
X and Y
‘Separate’
(%)
‘Close’
(%)
‘Together’
(%)
Total cells
analysed
17
20
30
25
30
27
77
13
18
3
17
17
17
0
70
63
67
58
53
57
23
30
40
30
100
30
30
30
775
Table 2. Signal disposition in dividing blood lymphocyte
interphase nuclei
Associated (%)
Chromosome
no.
3
7
8
13
17
21
Overall mean %
‘Separate’
(%)
‘Close’
(%)
‘Together’
(%)
Total cells
analysed
54
58
54
22
70
26
47
18
12
18
28
6
24
18
28
30
28
50
24
50
35
50
50
50
50
50
50
300
DISCUSSION
The primary aim of this investigation was to test the hypothesis that an association of homologous chromosomes accompanies cell differentiation using human Sertoli cells as an
example of a differentiated, non-dividing cell type and stimulated lymphocytes as an example of an actively dividing cell
type. Certainly ‘nuclear differentiation’ can be visualized in
these cell types at an ultrastructural level. Stimulated lymphocytes have large irregular shaped nuclei with some peripheral
heterochromatin against the nuclear envelope and a large
central nucleolus with many small fibrillar centres and a large
amount of granular component (Hozák et al., 1989; Fawcett
and Raviola, 1994). In contrast Sertoli cells have a nucleus
which is ellipsoidal in outline with one or two deep invaginations and a large central nucleolus with characteristic spatial
segregation of the nucleolar components (Fawcett and Raviola,
1994; Wachtler et al., 1992). The availability of whole chromosome painting probes now makes it possible to analyse the
positions occupied by any pair of human homologues within
the nucleus at any stage of cell division and differentiation. By
this means we have shown that there is a significant difference
Fig. 2. Adult human Sertoli cells. (A) Two nuclei each
showing a single signal for chromosome pair no. 7. (B) A
single signal for pair no. 13. (C) A single signal for pair
no. 21. (D) The X (red) and Y (green) chromosomes as
separated signals. The X occupies a peripheral position, the
Y is usually central.
in the distributions of the chromosomes in Sertoli cells and
lymphocytes. An examination of the data for individual chromosomes lends support to the idea of Leitch et al. (1994) that
homologue association is correlated either with frequency of
mitosis or to cellular differentiation in the most general sense.
In lymphocytes, a dividing cell type, there is no tendency
towards association for chromosomes 3, 7, 8 and 17. The acrocentric chromosome pairs 13 and 21 may behave slightly differently due to activity of the nucleolar organizing region. Previously, Sigmund et al. (1979) noted associations of these
chromosomes in lymphocyte metaphases. In addition, Mosgöller
et al. (1991) noted in dividing fibroblasts that there was a
tendency for larger chromosomes to be peripheral and smaller
chromosomes to be central on the mitotic plate: both observations might result in increased association of the smaller homologues. A lack of association of homologues in dividing cell types
has previously been noted in a number of cell types: amniotic
cells (Popp et al., 1990), fibroblasts (Emmerich et al., 1989) and
lymphocytes (Fergusson and Ward, 1992). In contrast, in Sertoli
cells there was a tendency towards association of the autosomal
776
A. C. Chandley, R. M. Speed and A. R. Leitch
pairs 3, 7, 8, 13, 17 and 21 whilst the opposite tendency was true
for the sex chromosomes. Other evidence for association of
homologous chromosomes in non-dividing cell types in
mammals exists for cells of the brain (Arnoldus et al., 1989, 1991;
Manuelidis and Borden, 1988; Borden and Manuelidis, 1988).
Sertoli cells are somatic cells in the mammalian testis which
provide support to the developing germ cells. From puberty
onwards, they persist as fully differentiated (non-dividing) cell
types, sperm production rates being closely related to Sertoli
cell numbers (Johnson et al., 1984). Does the association of
autosomal homologue pairs in Sertoli cells indicate a requirement for normal gene expression and cell regulation? If so,
could this pattern of chromosome behaviour be required for
normal gene expression in other non-dividing (differentiated)
cell types in man and other mammals? A link between centromere association and nucleolar activity in Sertoli cells of the
mouse has been demonstrated by Haaf et al. (1990). They
found that when centromeres were closely associated, in perinucleolar chromocentres, the nucleolus was transcriptionally
active: When not associated, it was inactive. The authors
concluded that centromere positioning plays an important role
in regulating Sertoli cell gene expression. Our present demonstration of autosomal homologue association in human Sertoli
cells covering the range of sizes across the genome, could also
indicate a relationship between chromosome positioning and
gene regulation. Studies to investigate this are planned.
In Drosophila, some clues regarding the need for a close association of homologues, and normal gene function have already
been found. Tartof and Henikoff (1991) have summarized a
general class of phenomena which they refer to as ‘trans-sensing
effects’, which share the common feature of one gene sensing the
presence of its homologue in trans. One good example cited
concerns polytene chromosome puffs. In such cases, a mutant site
will puff and accumulate mRNA when paired with its wild-type
homologue, but not when the chromosomes are asynapsed or
remain as paired mutant homozygotes. In this and several other
cited examples from Drosophila, gene expression becomes more
normal when homologues are paired. Tartof and Henikoff (1991)
view the trans-sensing process, both in diptera and in mammals,
as a pathway of interactions whose final physiological result is
appropriate gene expression. Interference at any point in the
pathway may cause dysfunction and, in humans, they believe that
any impediment to homologous chromosome pairing that is
required for normal gene expression, could have serious consequences for human health and development. A search for homologous pairing disruption in the Sertoli cells of selected infertile
males could prove of interest, for it might give clues to an underlying cause of spermatogenic disturbance. In the wider context,
an understanding might be gained of the phenomena underlying
normal gene expression in human tissues.
Dr R. Nichols is gratefully acknowledged for statistical advice and
assistance. The staff of the Photographic and Illustration Department,
MRC Human Genetics Unit, Edinburgh are thanked for the preparation of the Figures. Dr John Gosden is thanked for help with the
PRINS investigations.
REFERENCES
Arnoldus, E. P. J., Peters, A. C. B., Bots, G. T. A. M., Raap, A. K. and van
der Ploeg, M. (1989). Somatic pairing of chromosome 1 centromeres in
interphase nuclei of human cerebellum. Hum. Genet. 83, 231-234.
Arnoldus, E. P. J., Noordermeer, I. A., Peters, A. C. B., Raap, A. K. and van
der Ploeg, M. (1991). Interphase cytogenetics reveals somatic pairing of
chromosome 17 centromeres in normal human brain tissue, but no trisomy 7
or sex chromosome loss. Cytogenet. Cell Genet. 56, 214-216.
Bennett, M. D. (1984). Towards a general model for spatial law and order in
nuclear and karyotypic architecture. In Chromosomes Today, vol. 8 (ed. M.
D. Bennett, A. Gropp and U. Wolf), pp. 190-202. George Allen and Unwin.
Borden, J. and Manuelidis, L. (1988). Movement of the X chromosome in
epilepsy. Science 242, 1687-1691.
Chandley, A. C., Speed, R. M. and Ma, K. (1994). Meiotic chromosome
preparation. In Methods in Molecular Biology, vol. 29, Chromosome
Analysis Protocols (ed. J. R. Gosden), pp. 27-40. Humana Press, New Jersey.
Cooke, H. J. (1976). Repeated sequence specific to human males. Nature 262,
182-186.
Emmerich, P., Loos, P., Jauch, A., Hopman, A. H. N., Wiegart, J., Higgins,
M. J., White, B. N., Ploeg, M. van der, Cremer, C. and Cremer, T. (1989).
Double in situ hybridization in combination with digital image analysis: a
new approach to study interphase chromosome tomography. Exp. Cell Res.
181, 126-140.
Fawcett, D. W. and Raviola, E. (1994). Bloom and Fawcett - A Textbook of
Histology. Chapman and Hall, New York.
Fergusson, M. and Ward, D. C. (1992). Cell cycle dependent chromosomal
movement in pre-mitotic human T-lymphocyte nuclei. Chromosoma 101,
557-565.
Haaf, T., Steinlein, C. and Schmid, M. (1990). Nucleolar transcriptional
activity in mouse Sertoli cells is dependent on centromere arrangement. Exp.
Cell Res. 191, 157-160.
Hozák, P., Novák, J. I. and Smetana, K. (1989). Three-dimensional
reconstruction of nucleolus-organizer regions in PHA-stimulated human
lymphocytes. Biol. Cell 66, 225-233.
Johnson, L., Zane, R. S., Petty, C. S. and Neaves, W. B. (1984).
Quantification of the human Sertoli cell population: Its distribution, relation
to germ cell numbers, and age-related decline. Biol. Reprod. 31, 785-795.
Kofman-Alfaro, S., Speed, R. M., Boyle, S. and Chandley, A. C. (1994).
Condensation behaviour of the human X chromosome in male germ cells and
Sertoli cells examined by fluorescence in situ hybridization. Chrom. Res. 2,
439-444.
Leitch, A. R., Schwarzacher, T., Mosgöller, W., Bennett, M. D. and HeslopHarrison, J. S. (1991). Parental genomes are separated throughout the cell
cycle in a plant hybrid. Chromosoma 101, 206-213.
Leitch, A. R., Brown, J. K. M., Mosgöller, W., Schwarzacher, T. and
Heslop-Harrison, J. S. (1994). The spatial localization of homologous
chromosomes in human fibroblasts at mitosis. Hum. Genet. 93, 275-280.
Lewis, J. P., Tanke, H. J., Raap, A. K., Beverstock, G. C. and Kluin-Nelemans,
H. C. (1993). Somatic pairing of centromeres and short arms of chromosome 15
in the hematopoietic and lymphoid system. Hum. Genet. 92, 577-582.
Manuelidis, L. and Borden, J. (1988). Reproducible compartmentalization of
individual chromosome domains in human CNS cells revealed by in situ
hybridization and three-dimensional reconstruction. Chromosoma 96, 397-410.
Mosgöller, W., Leitch, A. R., Brown, J. K. M. and Heslop-Harrison, J. S.
(1991). Chromosome arrangements in human fibroblasts at mitosis. Hum.
Genet. 88, 27-33.
Popp, S., Scholl, H. P., Loos, P., Jauch, A., Stelzer, E., Cremer, C. and
Cremer, T. (1990). Distribution of chromosome 18 and X centric
heterochromatin in the interphase nucleus of cultured human cells. Exp. Cell
Res. 189, 1-12.
Sigmund, J., Schwarzacher, H. G. and Mikelsaar, A. V. (1979). Satellite
association frequency and numbers of nucleoli depend on cell cycle duration
and NOR activity: studies on first, second and third mitoses of lymphocyte
cultures. Hum. Genet. 50, 81-89.
Speed, R. M., Vogt, P., Köhler, M. R., Hargreave, T. B. and Chandley, A.
C. (1993). Chromatin condensation behaviour of the Y chromosome in the
human testis. Chromosoma 102, 421-427.
Tartof, K. D. and Henikoff, S. (1991). Trans-sensing effects from Drosophila
to Humans. Cell 65, 201-203.
Wachtler, F., Schöfer, C., Mosgöller, W., Weipoltshammer, K.,
Schwarzacher, H. G., Guichaoua, M., Hartung, M., Stahl, A., BergéLefranc, J. L., Gonzalez, I. and Sylvester, J. (1992). Human ribosomal
RNA gene repeats are localized in the dense fibrillar component of nucleoli:
light and electron microscopic in situ hybridization in human Sertoli cells.
Exp. Cell Res. 198, 135-143.
(Received 15 November 1995 - Accepted 8 January 1996)