Hum Genet (1986) 73 : 130-132
© Springer-Verlag 1986
The centromere index and relative length
of human high-resolution G-banded chromosomes
D . L . V a n Dyke 1, Maria J.Worsham 1, L. J. Fisher 2, and L. Weiss ~
1The Medical Genetics and Birth Defects Center and ZDepartment of Ophthalmology, Henry Ford Hospital, Detroit, MI 48202, USA
Summary. The relative length and centromere index were
compared in prometaphase and midmetaphase for each
human chromosome from five normal men. There were very
few differences between prometaphase and midmetaphase
chromosomes in these two parameters. Chromosome 7 had a
significantly different centromere index between prometa-
phase and midmetaphase, but no difference in relative length.
This was accounted for by significant differences in the rela-
tive length of both 7p and 7q between prometaphase and mid-
metaphase; 7p became relatively less condensed and 7q rela-
tively more condensed with progression from prometaphase
to midmetaphase. For chromosome 1, the short arm was sig-
nificantly longer than the long arm in both prometaphase and
midmetaphase, a finding that underscores the structural simi-
larity of this chromosome among the hominids.
Introduction
Idiograms of the human karyotype are useful in clinical cyto-
genetics, gene and phenotype mapping, comparative primate
cytogenetics, and in comparisons of mitotic chromosome mor-
phology with meiotic maps and linkage map distances. The
Paris Conference 1971 (1972) report provided idiograms and
measurements of the relative length and centromere index for
midmetaphase chromosomes. Standardized idiograms, but
not relative length and eentromere index measurements, of
the banding patterns for prometaphase, or high-resolution,
human chromosomes have been provided by the Standing
Committee on Human Cytogenetic Nomenclature (ISCN
1978, 1981) and by others (Francke 1981; Francke and Oliver
1978; Yunis 1980, 1981; Yunis et al. 1979). Yunis and cowork-
ers (1978) gave centromere index and other measurement
statistics for chromosomes 1-5 in four stages of chromosome
condensation. The idiograms usually depict the relative length
and centromere index of the chromosomes as unchanged from
late mitotic prophase through prometaphase to midmetaphase
and between meiosis and mitosis.
However, chromosome condensation is a dynamic process
which seems to proceed at different rates along the length of
each chromosome. Whereas Yunis and coworkers (1978,
1979) reported that only a few bands showed variability be-
tween midmetaphase and prometaphase, Francke and Oliver
(1978) reported that G-light regions condense more rapidly
than G-dark regions. Other workers reported that regions of
Offprint request to: D. L. Van Dyke, The Medical Genetics and Birth
Defects Center, Henry Ford Hospital, Detroit, MI 48202, USA
"preferential stretching" may be present on every chromo-
some (Richer et al. 1983; Drouin and Richer 1985). If G-
lighter chromosomes condense more quickly than G-darker
chromosomes, relative lengths should vary in different stages
of condensation. Also if G-lighter chromosome arms con-
dense more quickly than G-darker arms, centromere indexes
should vary. To determine whether these values vary signifi-
cantly between mitotic midmetaphase and prometaphase, we
compared the relative length and centromere index of mid-
metaphase and high-resolution, G-banded human chromo-
somes in five unrelated subjects.
Materials and methods
GTG-banded midmetaphase preparations (250.300 bands in
the haploid set) were made by standardized techniques. GTG-
banded prometaphase chromosome preparations (850-900
bands) were made from amethopterin-synchronized, phyto-
hemagglutinin-stimulated lymphocyte cultures using the
method of Yunis et al. (1978) with minor modifications (Babu
et al. 1984).
From each of five normal male volunteers of diverse ethnic
background (Arabic, Ashkenazi, Japanese, North European,
and South Indian), we selected 25 midmetaphase and 25 pro-
metaphase cells for measurement. In each cell at least one
member of each chromosome pair had good quality banding
with readily identifiable centromeres and chromosome ends.
Chromosome arm lengths were measured with a graphics
tablet and an Apple computer. For each chromosome, we cal-
culated the relative length (percentage of total autosomal
length) and centromeric index (length of short arm as a per-
centage of the whole chromosome length). A t least one copy
of each chromosome in each cell was measured, and whenever
possible both homologues were measured and the mean used
to calculate the relative length and centromere index. Using a
high resolution graphics tablet obviated the need to measure
only straight chromosomes, and provided more precise, accu-
rate measurements than would have been possible with calip-
ers or other instruments. The resolution of the graphics tablet
at the magnification used was less than 2 x 10- 2 gm per tablet
unit. Thus, the limiting factors in the final resolving power of
the measuring system were related mostly to the optics of the
photomicroscope and darkroom enlarger. The acrocentric
short arms were included in the present measurements, but
not the highly variable stalks or satellites. The 1978 ISCN
report midmetaphase measurements included the short arms,
stalks, and satellites.
 131

The relative lengths and centromere indexes were com- that one person, our subject E, showed the greatest differ-
pared by one-way analysis of variance with repeated measure- ences from the other four persons and this accounted for the
ments. Paired t-tests were employed for individual compari- significant interaction effect between subjects and chromo-
sons. A probability level of P < 0.02 was considered signifi- some condensation. The best explanation for this difference
cant, after employing the multiple testing correction method between subjects A - D and E seems to be a technical artifact;
of Bonferroni (Miller 1980). probably because the prometaphase cells selected for subject
E were at an earlier stage of metaphase (more than 900 bands
per haploid set) than those for subjects A - D (850-900 bands).
Results and discussion Analysis of variance for centromere index of the first four
persons (Table 1) showed significant differences between
For relative length (Table 1) we observed no significant differ- prometaphase and midmetaphase (P < 0.01) and significant
ences between prometaphase and midmetaphase (P > 0.31) interaction between individual chromosomes and chromo-
and no significant differences among the five subjects (P > some condensation, but there was no significant interaction
0.29). There was no indication of an interaction effect be- between subjects and chromosome condensation (P > 0.15).
tween subjects and the phase of condensation (P > 0.16), but To evaluate the interaction effect, the prometaphase and mid-
there was significant interaction between phase of condensa- metaphase centromere indexes of each chromosome were
tion and individual chromosomes (P < 0.01). To evaluate this compared by the t-test. Only chromosome 7 showed a signifi-
interaction effect, the prometaphase and midmetaphase rela- cant difference in centromere index between prometaphase
tive lengths of each chromosome were compared by the paired and midmetaphase (P < 0.01). From prometaphase to mid-
t-test, but no significant differences were observed. metaphase, the short arm of chromosome 7 became relatively
For centromere index there was a significant difference be- less condensed and the long arm became relatively more con-
tween prometaphase and midmetaphase (P < 0.01), but there densed. The relative length of 7p changed from 1.97 to 2.09
was also significant interaction between subjects and chromo- between pro- and midmetaphase (P < 0.036) and the relative
some condensation (P < 0.01) and between individual chro- length of 7q changed from 3.57 to 3.24 (P < 0.021). We also
mosomes and chromosome condensation (P < 0.01). Exami- compared the change in centromere index between prometa-
nation of the centromere indexes for each subject suggested phase and midmetaphase for each of the five subjects sepa-
rately, and observed no significant differences. Chromosome
7 did show a consistent, but nonsignificant, decrease in centro-
Table 1. Relative length and centromere index of human chromo- mere index from prometaphase to midmetaphase in all five
somes. The relative lengths and centromere indexes in midmetaphase subjects. Richer et al. (1983) noted that region 7q32.1---~qter
and high resolution chromosome preparations are compared. SD, was greatly stretched in their more elongated chromosomes, a
standard deviation. Any significant difference between midmetaphase
and high-resolution at overall alpha-level of 0.05 is indicated by an finding consistent with our observations.
asterisk Except for the chromosome 7 differences, we observed
little difference in relative length or centromere index be-
Chromo- Relative length Centromere index tween prometaphase and midmetaphase. Based on the ab-
some Mid- High- Mid- High- sence of systematic changes in relative length and centromere
number
metaphase resolution metaphase resolution index between prometaphase and midmetaphase, we found
Mean SD Mean 8D Mean SD Mean SD little to support the idea that condensation rates may differ
slightly for G-band light and dark regions of the karyotype
1 9.00 0.42 9.06 0.17 51.20 1.55 53.83 0.98
(Francke and Oliver 1978). However, comparison of whole
2 8.37 0.17 8.35 0.09 39.42 0.45 37.72 1.40
arm lengths is a less stringent test of that hypothesis than
3 6.96 0.20 6.81 0.14 46.36 0.95 46.73 3.53
would be comparison of region, band, or sub-band measure-
4 6.44 0.19 6.22 0.13 27.44 1.38 25.37 3.83
ments. Drouin and Richer (1985) described several regions of
5 6.16 0.14 6.02 0.05 26.59 0.95 26.53 2.55
differential condensation. They found that the R-bands which
6 5.95 0.18 5.90 0.10 35.69 3.49 37.77 1.05
varied most in length tended to be among the latest to repli-
7 5.38 0.18 5.53 0.05 38.61 0.61 35.59 0.39*
cate (6q23~q25, 7q36, and 11p15), and those which varied
8 4.77 0.16 4.83 0.12 34.16 0.91 32.37 1.31
least tended to be among the earliest to replicate (6p21, 6q27,
9 4.76 0.14 4.71 0.12 33.81 1.73 32.69 1.77
and 11p11.2). However, there were exceptions to this ten-
10 4.69 0.25 4.69 0.14 33.57 1.06 30.56 1.33
dency and the authors did not arrive at generalizations apply-
11 4.63 0.07 4.58 0.11 37.36 2.04 37.86 1.64
ing to all R-bands.
12 4.70 0.10 4.74 0.18 29.05 0.85 28.14 1.63
Because the centromere index of chromosome 1 was great-
13 3.41 0.04 3.36 0.10 10.74 0.91 9.50 0.40
er than 50 in both prometaphase and midmetaphase, we com-
14 3.32 0.07 3.34 0.09 11.61 0.32 9.40 0.47
pared the relative length of lp to that of l q by the t-test.
15 3.42 0.20 3.31 0.12 12.25 1.00 9.98 0.51
In both prometaphase and midmetaphase, the relative length
16 3.27 0.11 3.47 0.04 44.95 0.93 46.63 0.61
of lp was significantly greater than that of lq (P < 0.03 and
17 3.21 0.09 3.42 0.13 32.52 2.05 27.79 2.67
P < 0.001, respectively). This result compares favorably with
18 2.75 0.11 2.59 0.03 28.26 0.86 24.64 1.55
the centromere index measurement of human meiotic chromo-
19 2.63 0.07 2.78 0.10 45.55 2.18 48.32 0.71
some 1 (ISCN 1978). Chromosome 1 is homologous in humans
20 2.33 0.09 2.49 0.13 44.41 1.77 42.91 3.34
and in the great apes, but in the idiogram of the human karyo-
21 1.41 0.08 1.49 0.07 18.33 3.20 18.06 1.50
type the conventional orientation of chromosome 1 is up-side-
22 1.90 0.06 1.86 0.10 19.87 1.85 16.89 0.50
down with respect to that of the apes (Paris Conference 1971,
X 5.08 0.11 5.18 0.10 40.13 2.64 37.51 2.60
1975). When banding techniques were first introduced in
Y 2.02 0.30 1.72 0.22 27.84 2.07 23.49 2.87
1970 (Caspersson et al. 1970), the human chromosome 1 arm
132
with the block of paracentromeric h e t e r o c h r o m a t i n was
placed on the top, but its orientation was soon inverted,
apparently to be consistent with the prebanding orientation
(Schnedl 1971; Paris C o n f e r e n c e 1971, 1972).
Because polymorphism is greatest within the variable
regions on chromosomes 1, 9, 16, and the Y, we c o m p a r e d the
prometaphase and m i d m e t a p h a s e centromere index and rela-
tive length of these chromosomes among the five subjects
studied. The overall analysis of variance result was not signifi-
cant. Nevertheless, differences among individuals were signif-
icant (P < 0.02) for the Y c h r o m o s o m e total relative length in
p r o m e t a p h a s e and midmetaphase, and for the Y c h r o m o s o m e
c e n t r o m e r e index in midmetaphase. T h e r e were no differ-
ences among individuals for relative length of Yp in mid-
metaphase or p r o m e t a p h a s e , or for Yq in midmetaphase, but
the p r o m e t a p h a s e Yq differed significantly among subjects
(P < 0.04).
Acknowledgements. The authors thank Edward Peterson and Dr. Bar-
bara Tilley for statistical analyses, and Dr. Ramesh Babu for helpful
comments.
References
Babu VR, Van Dyke DL, Jackson CE (1984) Chromosome 20 dele-
tion in human multiple endocrine neoplasia types 2A and 2B: a
double blind study. Proc Natl Acad Sci USA 81 : 2525-2528
Caspersson T, Zech L, Johansson C (1970) Analysis of human meta-
phase chromosome set by aid of DNA-binding fluorescent agents.
Exp Cell Res 62: 490-492
Drouin R, Richer CL (1985) Analysis of high-resolution R-bands, ob-
tained by heat-denaturation and Giemsa staining, on human pro-
phase chromosomes. Can J Genet Cytol 27:83-91
Francke U (1981) High-resolution ideograms of trypsin-Giemsa
banded human chromosomes. Cytogenet Cell Genet 31 : 24-32
Francke U, Oliver O (1978) Quantitative analysis of high-resolution
trypsin-Giemsa bands on human prometaphase chromosomes.
Hum Genet 45 : 137-165
ISCN (1978) An international system for human cytogenetic nomen-
clature (1978). Birth Defects XIV:8
ISCN (1981) An international system for human cytogenetic nomen-
clature - high-resolution banding (1981). Birth Defects XVII:5
Miller RC (1980) Statistical inference, 2nd edn. Springer, New York,
pp 67-68
Paris Conference 1971 (1972) Standardization in human cytogenetics.
Birth Defects VIII : 7
Paris Conference 1971 (1975) Standardization in human cytogenetics
(Supplement). Birth Defects XI : 9
Richer CL, Drouin R, Murer-Orlando M, Jean P (1983) High-resolu-
tion idiogram of Giemsa R-banded human prophase chromo-
somes. Can J Genet Cytol 25 : 642-650
Schnedl W (1981) Analysis of the human karyotype using a reassocia-
tion technique. Chromosoma 34 : 448454
Yunis JJ (1980) Nomenclature for high resolution human chromo-
somes. Cancer Genet Cytogenet 2: 221-229
Yunis JJ (1981) Mid-prophase human chromosomes. The attainment
of 2000 bands. Hum Genet 56 : 293-298
Yunis JJ, Sawyer JR, Ball DW (1978) The characterization of high-
resolution G-banded chromosomes in man. Chromosoma 67 : 293-
307
Yunis JJ, Ball DW, Sawyer JR (1979) G-banding patterns of high-
resolution human chromosomes 6--22, X, and Y. Hum Genet
49 : 291-306
Received August 5, 1985 / Revised November 29, 1985