"Frontmatter". In: Plant Genomics and Proteomics

Adh1-F
Cinful
Ji
10 kb
FIGURE 1.5.
The structure of the Adh1-F region of maize, showing identified retro-
transposons. Only one gene is shown (Adh1-F), although more genes are present 
on this segment. The arrow above each element indicates its orientation. (Figure 
provided by Dr. J. Bennetzen.)

---

Centromeric DNA mediates chromosome attachment to the meiotic and
mitotic spindles and often forms dense heterochromatin. The Arabidopsis
genome sequence has identified the centromeric regions, which contain
numerous repetitive elements including retroelements, transposons,
microsatellites, and middle repetitive DNA. An unexpected observation was
that at least 47 expressed genes were encoded in the genetically defined cen-
tromeres of Arabidopsis (Copenhaver et al., 1999). The regions containing
these repeats also contain many more class I than class II elements (Figure
1.6). Because few centromeres, in fact only those from Arabidopsis and rice,
1 2
1. T
H E
S
T R U C T U R E O F
P
L A N T
G
E N O M E S
Mz
0.5
1.0
1.5
2.0
Mz
0.5
1.0
1.5
2.0
2.5
Predicted genes
Pseudogenes
Genes encoded
by mobile elements
Pseudogenes encoded
by mobile elements
Retroelements
Transposons
Characterized
centromeric repeats
180-bp repeats
Mitochondrial DNA
insertion
Chromosoma-spendillo
tandem repeat
Unannotated region
Expressed genes
A
B
*
**
*
* *
**
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
FIGURE 1.6.
Sequence features at CEN2 (A) and CEN4 (B). Central bars depict
annotated genomic sequence of indicated BAC clones; black, genetically defined 
centromeres; white, regions flanking the centromeres; //, gaps in physical maps.
Sequences corresponding to genes and repetitive features, filled boxes (above and
below the bars, respectively). (Reprinted with permission from Copenhaver et al.,
Science 286, 2468–2474. Copyright (1999) American Association for the Advancement
of Science.)

---

have been identified, the general structure of a centromere still must be
determined. Another unanswered question relates to the structure of the
kinetochore in comparison to the centromere. Will a kinetochore have an
attraction for transposons similar to that seen for the Arabidopsis centromere,
and so have a complex structure, or be a simpler stripped-down attachment
site, like that of yeast, that will make it easier to understand the essential
functions necessary for chromosome movement? There is evidence for 
conserved and variable domains among the centromere satellites from 
Arabidopsis populations (Hall et al., 2003).
The genes encoding the 18S, 5.8S, and 25S ribosomal RNAs are present
in tandem arrays of unit repeats in a recognizable chromosome structure, the
nucleolar organizer region (NOR). The repeat unit consists of the coding
sequences for each of these three RNAs as well as an internal transcribed
spacer region and an intergenic region (Figure 1.7). The number of repeat-
ing units varies between several hundred and over 20,000. Therefore, a plant
that has 20,000 copies of the ribosomal RNA genes has almost as much DNA
in this one tandemly arrayed family as Arabidopsis has in its whole genome.
The number of repeat units of these genes varies within a species and may
even vary within a plant (Rogers and Bendich, 1987). Even between maize
inbred lines the variation is more than twofold (Rivin et al., 1986). The vari-
ation in this gene family would account for a DNA difference of about 100
Mbp. Gymnosperms have a much longer repeat unit than angiosperms
(Cullis et al., 1988).
O
R I G I N O F
D N A V
A R I AT I O N
1 3
18
5.8
25
18
5.8
25
intergenic spacer region
FIGURE 1.7.
The repeat unit for the large ribosomal RNA genes.
P
ROCESSES
T
HAT
A
FFECT
G
ENOME
S
IZE
The genome can be extensively amplified by duplicating either part or all of
the genome through polyploidy. Polyploids have more than two complete
sets of chromosomes in their nuclei compared with the two that are found
in normal diploids. The rate of polyploidy in different groups is variable and
has been estimated as up to 80% in angiosperms, 95% in pteridophytes, but
relatively uncommon in gymnosperms. Polyploidy can arise in two differ-
ent ways (Figure 1.8). One of these is by doubling the chromosomes of a
single individual resulting in autopolyploidy. The other is by combining the
genomes from two closely related species. This latter event, which frequently
happens in a wide cross, results in the genomes of two different species resid-
ing in the same nucleus (allopolyploid). If the chromosomes from the two
genomes have diverged sufficiently so that the homologs from the two

---

species do not pair efficiently at meiosis, then the hybrid will be sterile.
However, a doubling of the chromosome number will result in a normal
meiosis and a new polyploid species will have been formed. Polyploids are
very frequent in the angiosperms, and most of the major crop species are
polyploids. These rounds of polyploidization are insufficient to account for
all of the increase in genome size seen in the angiosperms. To see an increase
of a thousandfold in the DNA content would require approximately 10
sequential rounds of doublings to have occurred. Octaploids seem to be the
largest frequently observed polyploids, only representing three sequential
doublings. However, the stonecrop is estimated to be about 80-ploid with
about 640 chromosomes (Leitch and Bennet, 1997). Despite this upper value,
the largest genomes are not the result of many rounds of whole genome 
doublings. 
Rather than the addition of a complete genome, various mechanisms can
result in the duplication of large regions of the genome. These mechanisms
include unequal recombination and nonreciprocal translocations. Both of
these mechanisms would result in one product having a loss of DNA while
the other has an increase. There would have to be a selective advantage for
the product that had a duplication in order for the genome to grow by this
method. Again, as pointed out for polyploidy, the number of rounds of
1 4
1. T
H E
S
T R U C T U R E O F
P
L A N T
G
E N O M E S
Diploid A
(n 
=
x 
=
4)
Autopolyploid
(2n 
=
4x 
=
16)
Allopolyploid
(2n 
=
4x 
=
22)
+
Diploid A
(n 
=
x 
=
4)
spontaneous doubling
Diploid A
(n 
=
x 
=
4)
Diploid B
(n 
=
x 
=
7)
+
FIGURE 1.8.
Mechanisms of polyploidization. (Reprinted from Trends in Pl. Sci. 2,
Leitch and Bennett, Polyploidy in Angiosperms, 470–476, Copyright (1997), with per-
mission from Elsevier.)

---

duplications needed to grow the genomes to the size that are seen are much
greater than would be supported by our current estimations of gene family
sizes. Therefore, other processes need to be operational apart from whole
and partial genome duplications. 
Most of the genome size increases in the grasses appear to be the result
of the amplification of retrotransposable element families. The retrotrans-
posons can increase their numbers within the genome because transposition
acts through an RNA intermediate. This therefore leaves the original DNA
copy in the genome, while placing additional copies elsewhere in the
genome. These elements should be acted on by natural selection so that their
continued expansion has led to their being named selfish or parasitic ele-
ments. The dispersal of rogue RNA polymerase III transcription products,
such as the Alu elements in humans and perhaps the expansion of the 5S
ribosomal RNA genes in flax, are demonstrations of this behavior. In all the
plants in which they have been investigated the LTR retrotransposons are
the biggest variable that can be related to genome size. These elements can
make up 60% or more of genomes like maize, wheat, and barley but less than
50% in rice and around 10% in Arabidopsis. The rice genome also contains
numerous inverted repeat transposable elements such as the MITES, which,
although numerous, are too short to have a large impact on the overall
genome size. 
The rounds of polyploidy and segmental duplications, along with trans-
posable element family amplification, all result in an increase in genome size
(Table 1.3). So are plant genomes destined to hold a one-way ticket to
O
R I G I N O F
D N A V
A R I AT I O N
1 5
TABLE 1.3. A H
YPOTHETICAL
C
ASE OF THE
E
FFECTS OF
G
ENOME
E
XPANSION
AND
C
ONTRACTION
Diploid chromosome
Event
number
2C DNA content
Ancestral genome
4
1 pg
Autopolyploidy
8
2 pg
Divergence, loss of DNA
8
1.9 pg
sequences
Retrotransposon explosion
8
2.3 pg
Alloploidy with species containing
20
4.3 pg
6 chromosomes as diploid 
number and 2 pg of DNA/2C
nucleus
Retrotransposon explosion
20
5.3 pg
Loss of chromosome pair
18
5.1 pg
(one of pairs containing 
ribosomal RNA genes)

---

“genomic obesity” (Bennetzen and Kellogg, 1997), or are there ways the
genome can be decreased? The loss of genes from polyploids has been
observed, sometimes at very high rates. These losses are associated with
deletions that are much smaller than the loss of whole chromosomes (Levy
and Feldman, 2002). 
Because much of the variation in genome size is associated with retro-
transposons, their removal could be an important factor in downsizing the
genome. Unequal recombination mechanisms can remove retrotransposon
sequences because the LTR regions are in a direct orientation and share a
very high degree of sequence homology. Therefore, if within a region of the
chromosome there were a number of insertions of related retrotransposon
sequences, then recombination between the two ends of this array would
result in a deletion of the array with the generation of a single LTR with no
other detectable associated LTR to define an intact element. The BARE-1
retroelement in barley demonstrates this phenomenon. The relative ratio of
solo LTRs to intact elements in barley and its wild relatives is consistent with
this model for retroelement copy number reduction (Bennetzen, 2002).