Content uploaded by Alexandre Hassanin
Author content
All content in this area was uploaded by Alexandre Hassanin
Content may be subject to copyright.
Fax +41 61 306 12 34
E-Mail karger@karger.ch
www.karger.com
Original Article
Cytogenet Genome Res 122:41–54 (2008)
DOI: 10.1159/000151315
Phylogenetic position of the saola (Pseudoryx
nghetinhensis) inferred from cytogenetic
analysis of eleven species of Bovidae
T.T. Nguyen a, b V.M. Aniskin c M. Gerbault-Seureau a H. Planton d
J.P. Renard
e B.X. Nguyen b A. Hassanin a V.T. Volobouev a
a UMR 5202 – Origine, Structure et Evolution de la Biodiversité, Département Systématique et Evolution,
Muséum National d’Histoire Naturelle, Paris (France)
b Institute of Biotechnology, Vietnamese Academy of Sciences & Technology, Hanoi (Vietnam)
c Severtsov Institute of Ecology and Evolution, Russian Academy of Sciences, Moscow (Russia)
d Chemin des Grandins, St Hilaire du Touvent (France)
e Institut National de la Recherche Agronomique, Jouy en Josas (France)
central Vietnam (Dung et al., 1993) to upward of 1500 indi-
viduals, of which 70–700 are distributed in Laos (Mallon
and Kingswood, 2001). The most recent estimates suggest
that less than 250 mature individuals survive in the wild
(IUCN, 2007).
The unusual combination of ‘caprine’ and ‘bovine’ mor-
phological characters make it difficult to determine the tax-
onomic position of saola with certainty, and two opposing
morphological hypotheses on its phylogenetic affinities
have been proposed. The anatomical study (cranial and
dental characteristics) of Thomas (1994) suggested that the
saola is attributable to the Antilopinae, and more precisely
to the tribe Caprini sensu lato which includes species close-
ly related to goat and sheep (Ropiquet and Hassanin, 2005).
Abstract.
Previous morphological and molecular analy-
ses failed to resolve the phylogenetic position of the criti-
cally endangered saola ( Pseudoryx nghetinhensis ) with re-
spect to its placement in Bovina (cattle, bison, and yak) or
Bubalina (Asian and African buffaloes). In the present
study, G- and C-banding, Ag-staining and FISH with 28S
and telomeric probes was undertaken for 17 bovid species.
An analysis of these data allowed us to identify 49 struc-
tural rearrangements that included autosomes, gonosomes
and 17 different NOR sites. The combined data set was sub-
jected to a cladistic analysis aimed at: (i) providing new in-
sights on phylogenetic relationships of the saola and other
species within the subfamily Bovinae, and (ii) testing the
suitability of different classes of chromosomal characters
for phylogenetic reconstruction of the family Bovidae. The
Request reprints from Dr. Vitaly Volobouev
UMR 5202 – Origine, Structure et Evolution de la Biodiversité
Département Systématique et Evolution
Muséum National d’Histoire Naturelle
16, rue Buffon, 75005 Par is cedex 05 (France)
telephone: +33 1 40 79 30 65; fax: +33 1 40 79 30 63
e-mail: vitaly@mnhn.fr
© 20 08 S. Karger AG, Bas el
142 4–8 581/0 8/122 1–00 41$2 4.5 0/0
Accessible online at:
www.karger.com/cgr
study revealed t hat nucleolar organizing regions (NORs) are
phylogenetically informative. It was shown that at least one,
or sometimes two of these characters punctuate divergenc-
es that include nodes that are the most basal in the tree, to
those that are the most recent. In this context, the shared
presence of three NORs in saola and species of Syncerus and
Bubalus strongly suggests the saola’s placement within the
subtribe Bubalina. This contrasts with Robertsonian rear-
rangements which are informative only at the generic level.
These findings suggest t hat NORs are an important and fre-
quently overlooked source of additional phylogenetic infor-
mation within the Bovidae that may also have applicability
at higher taxonomic levels, possibly even for Pecora.
Copyright © 2008 S. K arger AG, Basel
The work described in this paper was supported by grant from the French em-
bassy in Hanoi, Vietnam.
Accepted in revised form for publication by M. Schmid, 23 June 2008.
The saola ( Pseudoryx nghetinhensis ), until recently un-
known to science (Dung et al., 1993), joins several other bo-
vid species as being regarded as critically endangered by
IUCN (2007). Estimates of its population size va ry from sev-
eral hundreds within the restricted habitat in the north of
Cytogenet Genome Res 122:41–54 (2008)42
Other morphological studies, however, support its place-
ment within the Bovinae (cattle, buffaloes, spiral-horned
antelopes, nilgai, etc.) (Schaller and Rabinowitz, 1995; Ro-
bichaud, 1998).
In contrast, mitochondrial cytochrome b sequences al-
lied the saola with Bovinae and a provisional placement in
the tribe Boselaphini (the latter represented by the nilgai
Boselaphus tragocamelus ) (Dung et al., 1993). This initial
finding was further refined by the phylogenetic analyses of
combined mitochondrial and nuclear sequences which
similarly place this enigmatic species within the subfamily
Bovinae, but, importantly, within the tribe Bovini (Has-
sanin and Douzery, 1999b; Gatesy and Arctander, 2000;
Hassanin and Ropiquet, 2004). Hassanin and Douzery
(1999b) have defined three subtribes within Bovini: (i) Bo-
vina which incorporates species of the genera Bos and Bi-
son ; (ii) Bubalina which contains species of the genera Bub-
alus and Syncerus ; and (iii) the subtribe Pseudoryina, rep-
resented by P. nghetinh ensi s. The position of Pseudoryx
with respect to Bovina and Bubalina, however, remains un-
resolved (Hassanin and Douzery, 1999b; Hassanin and
Ropiquet, 2004).
In some mammalian taxa, particularly rodents, com-
parative chromosome banding studies have been extreme-
ly helpful in clarifying the systematic position of species
which do not display clear affinities when assessed by mor-
phological or/and molecular genetic characters (Taylor,
2000; Volobouev et al., 2002a, b, 2007; Dobigny et al.,
2003a). In the case of the Bovidae, however, in spite of the
large interspecific variation in diploid numbers and chro-
mosome morphology, chromosome banding has convinc-
ingly shown that this variation is largely attributable to
Robertsonian (Rb) rearrangements (Wurster and Be-
nirschke, 1968; Buckland and Evans, 1978; Gallagher and
Womack, 1992) most of which are species-specific (i.e. au-
tapomorphies), and this, combined with a scarcity of the
other types of structural changes, limits the usefulness of
these data in bovid phylogenetic reconstructions. In con-
trast, molecu lar cytogenetic stud ies of the bovid X chromo-
some have unambiguously demonstrated the importance
of changes involving this chromosome in resolving prob-
lematic evolutionary relationships (Ponce de Leon et al.,
1996; Robinson et al., 1996, 1997, 1998; Gallagher et al.,
1999). Additionally, preliminary data on the chromosomal
dist ribution of NORs among bov ids suggested that they too
may prove useful in a phylogenetic context (Gallagher et a l.,
1999).
Cytogenetic information on the saola is limited due to
accessibility of live animals (cultured fibroblasts exist from
a single female specimen worldwide). The first analysis
(Nguyen et al., 2005) provided the G-banded karyotype of
the species and the hypothesis that the 2n = 50 chromo-
somal complement resulted from a series of Rb fusions of
Bos -like ancestral acrocentric chromosomes. However, in-
terspecies comparisons were not performed. Subsequently,
G- and Q-banding for both saola and cattle was undertaken
by Ahrens et al. (2005) who provided for the identification
of chromosomes involved in Rb translocations using FISH
mapping of 32 BAC probes (i.e. one marker locus per chro-
mosome and two for pair 13 and X). No supportive banding
data was provided. Therefore, the questions on the extent of
whole-arm conservation between saola and other bovid
species remained unanswered.
In the present investigation cytogenetic data including
G- and C-banding, NOR location and number, and FISH
using ribosoma l 28S and telomeric probes obtained for saola
were compared for 16 bovid species. These were subjected
to cladistic analysis with three caprine species, Capra ibex ,
Rupicapra rupicapra and Ovis aries , as the outgroup. Our
findings provide further insights on the phylogenetic posi-
tion of Pseudoryx nghetinhensis within the subfamily Bovi-
nae, and demonstrate the utility of different classes of chro-
mosomal markers in phylogenetic reconstruction within
the family Bovidae.
Materials and methods
Samples
The list of the species studied and their main karyotype character-
istics are given in Table 1 . For saola, skin samples were collected from
a young female captured in the Bach Ma National Park, Thua Thien
Hue province, Vie tnam in 1994. The y were dipped in 1.5 M DMSO and
0.2 M sucrose and immediately frozen in liquid nitrogen to be used
subsequently for cell cu lture. The fibroblast cell cultures for other spe-
cies were available as cr yopreserved cells in the collection of the Mu-
séum National d’Histoire Naturelle (Paris).
Chromosome preparation and banding techniques
Chromosome preparations were obtained from fibroblast cell cul-
tures following standard protocols. G- and C-banding was carried out
as described by Seabright (1971) and Sumner (1972) respectively with
minor modifications. NORs were visualized by Ag-staining (Bloom
and Goodpasture, 1976). To ident ify NOR-bearing chromosomes, Ag-
staining was followed by Q-banding (Caspersson et al., 1972) of the
same metaphase spreads. At least 10 quality metaphases were ana lyzed
for each species.
N o m e n c l a t u r e
Chromosomes of the species studied as well as all chromosomal
re arr ang ements id ent ified we re e xpre sse d in acc orda nce to t he do mes -
tic cattle ( Bos taurus , BTA) standard nomenclature adopted by Inter-
national System for Chromosome Nomenclature of Domestic Bovids
(ISCNDB, 2000).
Fluorescent in situ hybridization
FISH using ribosomal 28S and telomeric probes was performed on
saola metaphase chromosomes as described by Gerbault-Serreau et al.
(2004). Hybridization results were examined and analyzed using a
Zeiss f luorescent microscope and the ISIS software package (Metasys-
tems, Alt lussheim, Germany). Metaphases were photographed with a
cooled CCD camera system Quanti x II (Photometrics, Tucson, AZ).
Cladistic analysis
The cladistic treatment of chromosomal data followed the princi-
ples detailed in Dobigny et al. (2004). Chromosomal changes were
identified on the basis of comparative analysis of banding patterns be-
tween a set of 14 bovi ne species (the ingroup) and th ree caprine specie s,
namely Capra ibex, Rupicapra rupicapra and Ovis aries that were used
as outgroup. In addition to structural rearrangements of both auto-
somes and gonosomes, we added available data on chromosomal loca-
tion of the NORs identified for most species by both FISH with 28S
ribosomal probe and Ag-staining. The chromosome changes and
NORs were used as characters and their presence/absence as character
Cytogenet Genome Res 122:41–54 (2008) 43
state coded ‘1’ or ‘0’ respectively. The missing data on the NORs distri-
bution for Taurotragus derbianus was coded as ‘?’. The data set was
analyzed using Maximum Parsimony (MP) in PAUP 3.1.1 (Swofford,
1993). Parsimony ana lyses were carried out with t he following options:
heuristic search with unlimited number of trees to be saved, swapping
with the tree-bisection-reconnection (TBR) algorithm, and random
addition of sequences with 100 replicates in order to increase the
chance of finding the most optimal tree(s). We performed successive
weighti ng using the retention i ndex to reweight each char acter (similar
results were obtained with consistency index or rescaled consistency
index; data not shown). Both Acctran and Deltran optimizations (ac-
celerated a nd delayed transformations, respectively) were compa red to
identify unambiguous apomorphies on the branches. To examine the
support for inferred relationships, bootstrap analyses were done using
1000 replicates with the closest stepwise addition option.
Results
Diploid numbers, chromosome morphology and the
karyotypes of the species analyzed correspond to earlier de-
scriptions (see Table 1 for reference). The only exceptions to
this are Taurotragus derbianus and Pseudoryx nghetinhensis
where our data suggested positional changes in the karyo-
types with respect to earlier descriptions (see Tables 1 and
2 for references).
Comparison of autosomal G-banding patterns
Genus Pseudoryx . The karyotype of the single female
saola analyzed (2n = 50, NFa = 58) comprises five pairs of
biarmed and 19 pairs of acrocentric autosomes; the two X
chromosomes are the largest acrocentric elements in the set
( Fig. 1 ). All the saola acrocentric chromosomes and the
chromosomal arms of its biarmed autosomes (except 5p)
correspond closely to their G-band homologues in karyo-
type of the cattle (see Table 2 for cattle equivalents involved
in the various rearrangements). The short arm of the pair 5
appears to be modified by paracentric inversion making
band-by-band comparisons difficult.
Genus Bos . Domestic cattle ( Bos taurus ) and banteng ( B.
javanicus ) have similar karyotypes with 2n = 60 consisting
of 29 pairs of acrocentric autosomes. The gaur ( B. frontalis )
has 2n = 58, the result of an Rb translocation involving an-
cestral chromosomes equivalent to BTA2 and 28. Of these
three species only the haploid set of the cattle is presented
as part of Fig. 1 . All these taxa possess similar submetacen-
tric X chromosomes.
Genus Bison . The G-banded karyotype of B. bonasus is
similar to that of the cattle and consequently not shown
herein.
Genus Bubalus . The swamp buffalo ( B. bubalis ) and the
anoa ( B. depressicornis ) share the same diploid number
Tab le 1.
Karyotype characterization of the studied species
Species, their common and abbreviated namesaSex 2nbNFacXdRearrangements identifiede
Bos taurus (Cattle), BTA (1) F 60 58 Sm –
Bos indicus (Zebu cattle), BIN (1)fM6058Sm–
Bos frontalis (Gaur), BFR (1) M 58 58 Sm t(2;28)
Bos javanicus (Banteng), BJA (1) F 60 58 Sm –
Bison bonasus (European bison), BBO (1) F 60 58 Sm –
Bison bison (American bison), BBI (1)fM6058Sm–
Bubalus bubalis (Swamp buffalo), BBU (1) F 48 56 A t(5;28;7), t(1;27), t(2;23), t(8;19), t(16;25)
Bubalus depressicornis (Lowland anoa), BDE (1) M 48 58 A t(1;27), t(2;23), t(8;19), t(5;28), t(11;20), t(17;25)
Bubalus mindorensis (Tamaraw), BMI (2)gM, F 46 56 A t(5;28;11), t(2;23), t(8;19), t(4;14), t(16;29)
Syncerus caffer (African buffalo), SCA (1)hM 52 58 A t(1;13), t(2;3), t(5;20), t(11;29)
Pseudoryx nghetinhensis (Saola), PNG (1) F 50 58 A t(1;10), t(8;13), t(6;19), t(4;18), t(11;12), inv12
Taurotragus oryx (Eland), TOR (1) F 32 56 St t(3;22;2), t(5;10), t(6;11), t(1;25), t(4;12), t(8;24), t(9;20), t(7;27),
t(15;16), t(18;19), t(14;26), t(21;23), t(17;28)
Taurotragus derbianus (Giant eland), TDE (1) M 31 56 St t(3;22;2), t(5;10), t(6;11), t(1;25), t(4;12), t(8;24), t rcp (9; 20dis)i,
t(7;27), t(15;16), t(18;19), t rcp (14;20prx;26)i, t(21;23), t(17;28)
Boselaphus tragocamelus (Nilgai), BTR (4)jM, F 46 56 Cm t(1;5), t(2;3), t(6;13), t(8;12), t(19;27), t(24;25)
Capra ibex (Alpine ibex), CIB (1) M 60 58 A t rcp (9;14)
Rupicapra rupicapra (Chamois), RRU (1) M 58 58 A t(1;3), t rcp (9;14)
Ovis aries (Domestic sheep), OAR (1)fM 54 58 A t(1;3), t(2;8), t(5;11), t rcp (9;14)
a The numbers in brackets denote the number of specimens examined.
b 2n: diploid number.
c Number of autosomal arms.
d Sm: submetacentric, A: acrocentric, St: subtelocentric and Cm: compound metacentric chromosome resulting from gonosome – autosome
translocation (see text).
e Numbered in accordance with cattle standard chromosome nomenclature (ISCNDB 2000).
f Gallagher et al. (1999).
g Tanaka et al. (2000).
h Gallagher and Womack (1992).
i See text for explanation.
j Gallagher et al. (1998).
Cytogenet Genome Res 122:41–54 (2008)44
(2n = 48) but differ in NFa, with 56 in the former, and 58
in the latter species. The reduction in diploid number is
attributable to five Rb translocations in the swamp buffalo
karyotype and six in the anoa, four of which are common
to both species: t (1;
27), (2; 23), (8; 19) and (5; 28). In B. buba-
lis , one Rb translocation (Rb 5;
28) was involved in a tandem
translocation with cattle chromosome 7 leading to the re-
duction of both 2n and NF ( Fig. 2 , Table 1 ). The X chromo-
somes of both species are acrocentric.
Genus Ta urot ragus . The karyotype of the female eland
( T. or y x ) studied corresponds to earlier descriptions (Buck-
land and Evans, 1978; O’Brien et al., 2006). The sex chromo-
some system in this species is X
1 X 1 X 2 X 2 /X 1 X 2 neoY as a re-
sult of the Y-autosome (BTA13) translocation, and thus the
females possess an even diploid number whereas the males
have one chromosome less. The male of giant eland ( T. der-
bianus ) analyzed in our study had 2n = 31 and an X
1 X 2 Y sex
chromosomes system and it was also heterozygous for a re-
ciprocal translocation involving pairs 7 (BTA9;
20) and 11
(BTA14;
26) in the eland karyotype ( Fig. 3 ). Given that T.
oryx and T. derbianus h ave pre viousl y been repor ted to pos -
sess similar karyotypes (O’Brien et al., 2006), it is possible
that the giant eland here is a carrier of spontaneous recipro-
cal translocation. Further studies are needed to clarify the
origin of this rearrangement and its frequency in natural
populations of T. derb ianu s .
G e n e r a Capra and Rupicapra . The alpine ibex ( C. ibex )
and chamois ( R. rupicapra ), with 2n = 60 and 58 respec-
tively, differ from each other by one Rb translocation in-
volving cattle equivalents BTA1 and BTA3 (karyotypes not
shown). In addition, both are characterized by reciprocal
translocations between cattle equivalents 9 and 14, a rear-
rangement characteristic of numerous non-bovine species.
The X chromosomes are acrocentric.
C - b a n d i n g
Blocks of constitutive heterochromatin were detected in
pericentromeric regions of all acrocentric chromosomes in
all species studied ( Fig. 4 ). However, these varied in size
from pair to pair and from species to species and are thus
difficult to quantify. Compared to acrocentrics, the amount
of C heterochromatin on the biarmed autosomes and sub-
metacentric X chromosomes in Bos and Bison is reduced,
and in some species totally absent. Comparisons of the C-
banding patterns failed to yield clear-cut interspecies differ-
ences.
Tab le 2.
Comparison of G-banding patterns between 11 species studied
BTA BFR BJA BBO BBU BDE PNG TOR TDE CIB RRU
1 1 1 1 2q 1q 1q 4q 4q 1 1q
2 2q 2 2 3q 2q 6 1q prx 1q prx 2 2
3 3336791p1p31p
4 4 4 4 7 8 4q 5q 5q 4 4
5 5 5 5 1p 4q 7 2q 2q 5 5
6 6 6 6 8 9 3q 3q 3q 6 6
7 7 7 7 1q dis 10 8 8q 8q 7 7
8 8 8 8 4q 3q 2q 6q 6q 8 8
9999911107q7q
a9b9b
10 10 10 10 10 12 1p 2p 2p 10 10
11 11 11 11 11 5q 5q 3p 3p 11 11
12 12 12 12 12 13 inv 5p 5p 5p 12 12
13 13 13 13 13 14 2p cc13 13
14 14 14 14 14 15 15 11q 11qa14b14b
15 15 15 15 15 16 14 9q 9q 15 15
16 16 16 16 5q 17 12 9p 9p 16 16
17 17 17 17 16 6q 17 13q 13q 17 17
18 18 18 18 17 18 4p 10p 10p 18 18
19 19 19 19 4p 3p 3p 10q 10q 19 19
20 20 20 20 18 5p 11 7p 7pa20 20
21 21 21 21 19 19 16 12q 12q 21 21
22 22 22 22 20 20 18 1q dis 1q dis 22 22
23 23 23 23 3p 2p 19 12p 12p 23 23
24 24 24 24 21 21 22 6p 6p 24 24
25 25 25 25 5p 6p 13 4p 4p 25 25
26 26 26 26 22 22 21 11p 11pa26 26
27 27 27 27 2p 1p 23 8p 8p 27 27
28 2p 28 28 1q prx 4p 20 13p 13p 28 28
29 28 29 29 23 23 24 15 15 29 29
a Involved in another reciprocal translocation (see text).
b Involved in reciprocal translocation (see text).
c Translocated onto Y chromosome (see Fig. 3).
Cytogenet Genome Res 122:41–54 (2008) 45
X chromosomes
There are three main types of the X chromosomes pres-
ent among the species studied (one submetacentric in mor-
phology and two acrocentric types). A submetacentric X is
characteristic of all species of the genera Bos and Bison . Of
the two acrocentric X chromosome variants, one type is
found in Capra and Rupicapra , whereas the second is shared
by saola, swamp buffalo, anoa and, except minor differenc-
es, also by the two eland species ( Fig. 5 ). Data from com-
parative banding and the integration of molecular cytoge-
netic studies of the X chromosomes in bovids (Ponce de
Leon et al., 1996; Robinson et al., 1996, 1998; Gallagher et
al., 1999) allowed us to identify the rearrangements ( Fig. 6 )
that have shaped the morphological diversity of the X chro-
mosome and the sex chromosome systems in modern Bo-
vidae.
Fig. 1.
C omparison of G-band ed chromosomes betwe en saola Pseu-
doryx nghetinhensis (numbered below) and cattle Bos taurus .
10
13
19 18
1
12 3 4 5
6
8
12
411
2573
9
20 16 25 15 14
21 17 22 23 28
26 24 27 29 X
67 8910
11 12 13 14 15
16 17 18 19 20
21 22 23 24 X
4
10
123
6p
17
78911 12
5q 13 14 15 16
6q 18 5p 19 20
21 22 23 X
12345
678910
11 12 13 14 15
16 17 18 19 20
21 22 23
X
1 2 3 4 5
86 9 10
12 13 14
TDE TOR
TDE
TOR
11 7 X1X1 X2X2X1X2Y
TOR TDE
20
9
26
14 X13
3
22
2
10
5
11
6
26
1
12
4
24
8
27
7
16
15
18
19
23
21
28
17 29
Fig. 2.
Comparison of G-banded chromosomes between swamp
buffalo Bubalus bubalis (numbered below) and lowland anoa Bubalus
depressicornis .
Fig. 3.
C omparis on of G-banded chromosome s between eland Tau-
rotragus oryx (left) and giant eland Taurotragus derbianus (right), and
their correspondence to the cattle chromosomes. Insert: chromosome
pairs 7 and 11 involved in translocation.
Cytogenet Genome Res 122:41–54 (2008)46
abcd
efgh
ijk
4
Cytogenet Genome Res 122:41–54 (2008) 47
N O R s
The number and chromosomal location of the NORs
was determined for P. ngh etinhensi s and B. depressicornis
( Fig. 7 ) since these species were not included in Gallagher
et al. (1999). With Ag-staining, P. ngh etinh ensi s shows sev-
en NOR bearing chromosome pairs, the highest number
recorded in Bovidae. All the NORs are located in telomer-
ic regions and were present on both homologues of pairs
2, 4, 18 and 21 (corresponding to BTA8, 18, 22 and 26), and
only one homologue of pairs 8, 15 and 19 (corresponding
to BTA7, 14 and 23). NORs are also telomerically located
in the anoa being detected on one homologue of six auto-
somal pairs corresponding to BTA3, 14, 19, 23, 25 and 26
( Fig. 7 ).
BTA BFR BJA BBO PNG BBU BDE TOR TDE RRU CIB
1
2
3
4
5
6
1
2
3
4
5
6
1
4
5
6
1
2
3
4
5
6
6
5
3
2
4
1
*
III III IV V
23
2
3
6
7
15
22
8
4
21
15
19
4
2
2
18
18
21
ab
Fig. 7.
Si lver stai ning of sao la P. n ghetinhen sis (
a
) and anoa B. depressicornis (
b
) metapha se spreads. Arrows show the
NOR-bearing chromosomes.
Fig. 5.
Three main types of the X chromosome among the species
studied. Black dots indicate the centromere position (see text for de-
tails).
Fig. 6.
Schema of the X chromosome transformation in bovids fol-
lowing molecular cytogenetic data by Robinson et al. (1998) and Gal-
lagher et a l. (1999). I – Caprini type, II – Bubalus type, III – Bos type,
IV – Tr age l aph u s ty pe and V – Boselaphus ty pe. Numbers on the chro-
mosomes show position of the six BACs mapped. I ]
II – 3 inversions,
II ] III – centromere transposition, III ] IV – gonosomes-BTA13
translocation, III ]
V – gonosomes-BTA14 translo cation. * Tw o BACs
in eland (IV) are co-localized (for details see Gallagher et al., 1999).
Fig. 4.
C-banded karyotypes of (
a
) Bos taurus , (
b
) Bos frontalis ,
(
c
) Bos javanicus , (
d
) Bison bonasus , (
e
) Bubalus bubalis , (
f
) Pseudo-
ryx nghetinhensis , (g) Bubalus depressicornis , (
h
) Taurotragus oryx ,
(i) Taurotragus derbianus , ( j) Capra ibex and (k) Rupicapra rupicapra .
5
6
Cytogenet Genome Res 122:41–54 (2008)48
F I S H a n a l y s i s
The 28S ribosomal probe revealed the same number and
chromosomal location of 28S genes as those detected by sil-
ver staining ( Fig. 8 a, b). This means that saola was hetero-
zygous for the presence/absence of NORs on three autoso-
mal pairs corresponding to BTA7, 14 and 23.
FISH using the (TTAGGG)
n telomeric probe produced
signals exclusively at the terminal ends of all chromosomes
( Fig. 8 c, d). The intensity of hybridization signals was con-
stant and each chromosome general ly exhibited four signals
confirming the telomeric pattern previously found in most
bovid species (Meyne et al., 1990; de la Seña et al., 1995;
Tanaka et al., 2000).
Chromosomal phylogeny
We complemented our comparative banding analysis of
the 11 bovid species presented here through the inclusion of
published information from six additional taxa: Bos indicus ,
Bison bison , Boselaphus tragocamelus , Syncerus caffer , Bu-
balus mindorensis and Ovis aries (see Table 1 for references).
Using three caprine species as outgroup, we identified all
putative euchromatic autosomal homologies and thus all
structural rearrangements that occurred during the evolu-
tion of the ingroup species. These comprise 39 Robertso-
nian rearrangements, three tandem and two reciprocal
translocations and one paracentric inversion ( Table 1 ).
Additionally, we included four rearrangements of the sex
chromosomes and 17 additional characters based on chro-
ab
cd
Fig. 8.
Representative FISH images of ribosomal 28S (
a
) and telomeric (
c
) probes on s aola metaphases. Chromosomes
were identified using reverse PI banding (
b
and
d
). Arrows indicate sites of probe hybridization.
Cytogenet Genome Res 122:41–54 (2008) 49
mosomal distribution of the NORs (see Table 3 for refer-
ences).
The treatment of the chromosomal character matrix (see
Appendix) by the MP analysis resulted in three equally par-
simonious trees of 79 steps (CI = 0.892, RI = 0.883). We per-
formed successive weighting using the retention index to
reweight each character and obtained the single tree pre-
sented in Fig. 9 a. We retrieved the same topology when us-
ing consistency index or rescaled consistency index for suc-
cessive weighting. The bootstrap (BP) analysis showed the
presence of five robust nodes: (1) the separation between
Bovinae and Caprini (BP = 95%); (2) the grouping of Bos and
Bison species (subtribe Bovina, BP = 84%); (3) the mono-
phyly of Bubalus (BP = 90%); (4) the sister-group relation-
ships between B. bubalis and B. depressicornis (BP = 86%);
and (5) the monophyly of Tau ro trag us (BP = 100%). All oth-
er nodes are supported by BP values ! 7 0 % . W e e x a m i n e d
distribution of the parsimony-informative characters cate-
gory by category to ga in a further insight on their relevance
to the tree topology, as well as to the cladogenetic process.
Of 66 chromosomal characters analyzed, 34 are phylo-
genetically informative. These comprise 16 Rb rearrange-
ments, 12 NORs, one tandem and one reciprocal transloca-
tion involving autosomes, and four rearrangements of the
sex chromosomes. Among 16 Rb rearrangements, 10 were
shared by two species of Taurot ragus (characters 29 to 33, 35
to 37, 39 and 40), four by three species of Bubalus (characters
2, 5, 6 and 7), one by Boselaphus and Syncerus (ch arac ter 14)
and one by Tauro trag us and Bubalus mindorensis (c harac ter
31). When mapped to the tree 12.5% of the Rb rearrange-
ments are homoplasic characters.
The analysis of the NORs, the second largest category of
characters (n = 17) used, showed that 12 (70.6%) are phylo-
genetically informative, some of which display remarkable
stability being preserved in many taxa since the Bovidae/
Cervidae split (characters 47 and 48) (see Gallagher et al.,
1999 for data on Cervidae) or the Bovinae/Antilopinae split
(character 46). Others mark more recent events (characters
53 and 58) and sometimes are characteristic of individual
species thus bei ng autapomorphies for these linea ges (50, 51,
52, 55 and 56). Although the loss of ancient NORs and the
emergence of new chromosomal sites bearing NORs often
occur in concert (i.e., chars. 46, 48, 53, 58, 61), their occur-
rence in the taxa belonging to unrelated lineages (i.e., chars.
54 and 62) was frequently found. As a consequence, NOR
characters exhibit higher levels of homoplasy than do Rb
translocations (53.8% vs. 12.5%). In contrast to autosomal
characters, all changes of the X chromosome appeared to be
strong phylogenetic signals. No instances of homoplasy
were detected.
84
52
53
51
95
58
90
57
86
100
63
53 64
49 60
46 58
61
6757
63
48
45
1417 18 19 20 21 47 54 62 66
22232425 26 27 47 50 51 54 56 59 60 62
25
4111231 60
910 54 58 59 62
38
2829303132 33353637 3940 65
3438 52 55
41
4344
42
48 62
1
13141516
Robertsonian translocation NORs
Reciprocal translocation
Inversions (X)
Tandem translocation
Transposition (X)
Autosomal inversion
Gonosome-autosome translocation
Ambiguous location
Bos taurus taurus
Bos taurus indicus
Bos javanicus
Bos frontalis
Bison bison
Bison bonasus
Syncer us caer
Bubalus depressicornis
Bubalus bubalis
Bubalus mindorensis
Pseudoryx nghetinhensis
Boselaphus tragocamelus
Tau rot rag us oryx
Taurotragus derbianus
Rupicapra rupicapra
Ovis aries
Capra ibex
2 4 6 8 10 12 14 16 18 20Mya
ab
Fig. 9.
Phylogenetic relationships between 17 species of Bovidae
based on chromosomal (
a
) and molecular (
b
) data. The tree on the left
(
a
) is the strict consensus tree of three equally parsimonious trees of 79
steps (consistency index = 0.892), obtained from the treatment of the
chromosomal characters matrix (66 characters) by maximum parsi-
mony (MP). The symbols on the branches represent chromosomal re-
arrangements and distribution of NORs. The numbers below the sym-
bols correspond to chromosomal character numbers (see Appendix),
and the empty symbols correspond to homoplasic characters. The
numbers in grey circles are % bootstrap values (1000 replicates). The
tree on the right (
b
) is obtained from the analyses of complete mito-
chondrial cytochrome b (1143 bp), 12S rRNA (956 bp) genes, non-cod-
ing regions from the nuclear genes for aromatase cytochrome P-450
(199 bp) and lactoferrin (338 bp) (Hassanin and Douzery, 1999b), and
from sequences of the promoter of the lactoferrin ( Lf ), and two mito-
chondrial genes, i.e., the cytochrome b and the subunit II of cyto-
chrome c oxidase ( CO2 ) (Hass anin and Ropiquet, 2 004) using Bayesian
and MP methods. The data for Bubalus species are from Tanaka et al.
(1996). The molecular time scale presented is calculated according to
Hassanin and Ropiquet (2004).
Cytogenet Genome Res 122:41–54 (2008)50
Discussion
Chromosomal changes in evolution of the family Bovidae
Comparative banding analysis involving 17 bovid taxa
revealed extensive inter-species monobrachial homologies,
a conclusion reached in earlier studies (Buckland and Ev-
ans, 1978; Gallagher and Womack, 1992; Gallagher et al.,
1999). Indeed, except for some rare chromosomal changes
(see below), the karyotypes of a ll bovine species stud ied may
be easily derived from that of a Bos -like ancestor by means
of numerous Rb and fewer tandem translocations (36 vs. 3),
i.e., rearrangements that have not disrupted ancient synte-
nies. Among Robertsonian rearrangements, shared Rb
combinations may occur in closely related as well as phylo-
genetically distant species. Thus, in the absence of rigorous
cladistic analysis it is difficult to decide which of these char-
acters are synapomorphic and which are homoplasic (i.e.,
Rb (2;
3) shared by Syncerus and Boselaphus , or Rb (1; 25)
which is shared by two species of eland and the tamaraw).
The Rb (1;
10) translocation which appears as a derived char-
acter in the saola has also been identified in two species of
gazelles ( G. dorcas and G. gazella , tribe Antilopini) (Vassart
et al., 1995), and two species of Damaliscus , D. lunatus and
D. pygargus , tribe Alcelaphini (Kumamoto et al., 1996). Yet
more confusing is the shared presence of the Rb (1;
25) in
tamaraw (Bovini), eland (Tragelaphini), gazelle (Antilopi-
ni) and goral (Caprini). We conclude that Rb rearrange-
ments may be useful phylogenetic characters but, in many
Tab le 3.
Chromosome location of the NORs in the species of the families Bovidae and Cervidae
Species NOR-bearing chromosomesaReferences
Bovidae
Bos taurus 2 3 4 11 25 28 Gallagher et al. 1999
Bos frontalis 2 3 4 11 25 Ibid
Bos javanicus 2 3 4 11 25 Ibid
Bos indicus 2 3 4 11 25 Ibid
Bison bison 2 3 4 11 25 28 Ibid
Bison bonasus 2 3 11 25 28 Ibid
Syncerus caffer 3 22 25 28 Ibid
Bubalus bubalis 3 19 22 25 26 28 Ibid
Bubalus depressicornis 3 14 19 23 25 26 This study
Bubalus mindorensis 3 19 22 26 28 Tanaka et al. 2000
Pseudoryx nghetinhensis 7 8 14 18 22 23 26 This study
Boselaphus tragocamelus 2 5 14 Gallagher et al. 1998, 1999
Tautrotragus oryx 2 3 5 10 16 28 Gallagher et al. 1999
Ovis aries 2 3 4 5 28 Ibid
Capra ibex 2 3 4 5 28 Mayr et al. 1987
Rupicapra rupicapra 2 3 4 5 28 Ibid
Cervidae
Cervus nippon 3 4 Gallagher et al. 1999
Odocoileus virginianus 3 4 Gallagher et al. 1999
a Numbered in accordance to cattle standard chromosome nomenclature (ISCNDB 2000).
cases, they may be homoplasic and thus their contribution
to defining cladogenetic events in the family Bovidae may
be rather modest. By not taking their rather high indepen-
dent occurrence into account, these chromosomal changes
can lead to misinterpretation of phylogenetic relationships
in the absence of appropriate analysis, as was previously
stated by Robinson et al. (1997).
Tandem translocations are well known for their severe
effect on hybrid fertility and viability which explains the
rarity of this chromosomal mutation in karyotype evolu-
tion of mammals (King, 1993). However, as soon as this re-
arrangement is fixed in a population it may lead to rapid
reproductive isolation (Taylor, 2000). In our study one tan-
dem translocation (BTA3;22;2) was identif ied in two species
of eland and one (BTA7;5;28) in the swamp buffalo ( Buba-
lus bubalis ). It is noteworthy that chromosome 1 of the tam-
araw ( B. mindorensis ) results from a tandem translocation
involving Rb chromosome (BTA5;28; Tanaka et al., 2000)
which is similarly present in the swamp buffalo, but which
is fused with BTA11 in the former and BTA7 in the latter
species. However, Tanaka et al.’s interpretation is likely er-
roneous since the same chromosome was matched twice –
once as chromosome R6, and then as chromosome R12 of
river buffalo; in the latter case it was presumably homolo-
gous to the tamaraw’s chromosome 1q ( Fig. 3 in Tanaka et
al., 2000). Similar errors occur elsewhere in their compari-
son and it seems that this disagreement results from a par-
tial lack of correspondence in chromosome numbering us-
ing G- and R-banded chromosome nomenclature, and the
Cytogenet Genome Res 122:41–54 (2008) 51
mismatching of at least two chromosomal pairs in the Tana-
ka et al. (2000) publication. Should this hold, the tandem
translocation (BTA7;5;28) unites Bubalus bubalis and B.
mindorensis rather than Bubalus bubalis with B. depressi-
cornis (as presented in Fig. 9 a), a finding which is consistent
with the molecular phylogenetic analysis ( Fig. 9 b).
Despite an elevated occurrence of de novo autosomal re-
ciprocal translocation in man (Rousseaux et al., 1995) and
domestic animals (Pinton et al., 1998), usually associated
with the production of aneuploid gametes and sometimes
leading to full sterility in mammals, this kind of rearrange-
ment is rarely detected in wild mammalian species (King,
1993; Searle, 1993 and references therein). One of the few
well documented cases involves Arvicanthis . These rodent
species show de novo autosomal reciprocal translocation at
the node marki ng the divergence of the two main evolution-
ary lineages (Volobouev et al., 20 02a). Similarly, the recipro-
cal translocation (BTA9;14) appears to be an important
event in the Bovidae where, together with the changes in X
chromosome morphology, it marks the basal split between
the subfamilies Bovinae and Antilopinae (species within
these groups are characterized by ‘bovine’ and ‘caprine’ X
chromosomes, and the presence/absence of the BTA9;14 re-
ciprocal translocation) (Buckland and Evans, 1978; Galla-
gher and Womack, 1992; Gallagher et al., 1999). Further-
more, the detection of a paracentric inversion in the saola is
noteworthy as this type of rearrangement has not previous-
ly been detected in Bovidae. This finding needs to be con-
firmed by analysis of more animals to exclude the possibil-
ity of mutation occurrence during cell culturing.
In contrast to the remarkable conservation of autosomal
banding patterns, three structural variants of the X chro-
mosome were detected among bovids studied supporting
previous findings (Buckland and Evans, 1978; Robinson et
al ., 1997, 1998; Gal lagher et al ., 1999). Import antly, however,
several bovid lineages show compound sex chromosomes
resulting from gonosome-autosome translocations (see
O’Brien et al., 2006 for references) as typified by the X
1 X 2 Y
eland system which is similarly shared by bongo Tragela-
phus eurycerus and lowland nyala T. anga s ii . After translo-
cation of the same autosome onto X, the sex chromosome
system evolved into neoXneoX/neoXneoY system as such
found in the lesser kudu (all three species from the tribe
Tragelaphini) (O’Brien et al., 2006) as well as in the nilgai
Boselaphus tragocamelus of the tribe Boselaphini, although
in this case another autosome (BTA14) is involved in trans-
location (Gallagher et al., 1998). All modifications of the X
chromosome, especially X-autosome translocations, are po-
tentially powerful reproductive isolating mechanisms due
to their impact on sex determination and/or on the X-inac-
tivation process (King, 1993).
NORs are a very different category of genetic marker
from the structural rearrangements considered above. In
brief, the evolutionary dynamics of NOR variation involve
changes in rDNA copy number (amplification – deletion)
and the chromosomal location of these simple multigene
families. In addition to their main function – production of
ribosomes – these genes are also involved in other impor-
tant cellular activities such as regulation of rRNA transcrip-
tion, formation of microtubule-associated proteins and nu-
cleolar cortical skeletal proteins (Sumner, 1990, 2003 and
references therein). The changes in NOR number and/or
chromosomal location are not known to directly influence
the reproductive performance and viability of their carriers.
Most mammalian species possess 1 to 5 chromosome pairs
that bear rDNA loci. There are, however, notable exceptions
to this. For example, most bat species of the genus Myotis
possess 1 to 4 NOR bearing chromosomal pairs but this ex-
tends to 14 chromosomal pairs in M. myotis (Volleth, 1987).
Although 1 to 2 of these pairs might be shared by conge-
neric species, as well as the species from the evolutionarily
close Eptesicus , Nyctalus and Vespertilio , no clear phyloge-
netic inference could be drawn from analysis of NOR chro-
mosomal distribution in the above bat taxa (Volleth, 1987).
Likewise, NOR distribution in t he gerbils ( Taterillus , Dobig-
ny et al., 2003b and references therein) was similarly phylo-
genetically uninformative. In other words, the lack of NOR
phylogenetic context spans species whose karyotype evolu-
tion is markedly different. Bats are highly conserved and
usually the species differ from each other by a few, easily
detectable changes of G-banding patterns, whereas the ger-
bil species show extensive and rapid karyotypic repattern-
ing. In contrast, however, there are examples of surprising-
ly long-term conservation of NOR chromosomal locations.
For example, FISH analysis of severa l 18S+28S and 5S rRNA
genes revealed their retention on homeologous chromo-
somes or chromosomal segments in genomes of babirusa
(Suidae) and collared peccary (Tayassuidae), taxa that di-
verged at least 35 Myrs ago and which have undergone ex-
tensive subsequent karyotype repatterning (Zijlstra et al.,
1997; Bosma et al., 2004). Furthermore, two NOR bearing
chromosomes found in cervids are shared by many bovid
species (Gallagher et a l., 1999), families that diverged about
28 Myrs ago (Hassanin and Douzery, 2003).
As it was discussed in detail by Dobigny et al. (2004), the
use of NORs to determine phylogenetic relationships should
be done with caution. Indeed the inability of Ag-staining to
detect transcriptionally non-active rRNA genes and the
nonspecific binding of silver nitrate with chromatin (Sum-
ner, 1990; Dobigny et al., 2002) may lead to errors in estima-
tion of NOR number and/or their chromosomal location.
Another problem concerning their identification is related
to the high variability of rRNA copy numbers within indi-
viduals and among species (Sumner, 1990, 2003; Gallagher
et al., 1999). This variability can impact on the detection of
low-copy rDNA loci due to limited sensitivity of FISH.
These d ata make it cle ar t hat i n ad dit ion to tec hnic al lim ita -
tions, a n appropriate sample size is crucia l in accurately de-
termining the number of rDNA loci and their chromosom-
al location within a particular species.
Although Gallagher et al. (1999) were the first to suggest
that the chromosomal distribution of NORs among Bovidae
may be phylogenetically useful, their data were not subject-
ed to a rigorous phylogenetic analysis. This is supported by
our cladistic analysis of the chromosomal distribution of
NORs in 16 bovid species. In fact, 70.6% of NORs are phy-
Cytogenet Genome Res 122:41–54 (2008)52
logenetically informative. We show that one to two of these
characters mark practically all divergence events from the
most basal of nodes to the most recent. This differs mark-
ed ly fr om R obe rt soni an rearr an geme nts wh ich app ea r to b e
variable only at a generic level ( Fig. 9 ).
The finding of NOR heterozygosity (i.e. only one chro-
mosome of a pair bears an NOR) on three chromosomal
pairs in saola mimics an earlier observation of double het-
erozygosity in Bos indicus (Gallagher et al., 1999). This
would seem to suggest that difference in the copy number
of 28S genes among bovid kar yotypes is not an unusua l phe-
nomenon. Given that the NOR data for our study were ob-
tained from single specimen by species (with the exception
of nilgai and tamaraw), it is highly probable that at least
some NOR sites were overlooked, and this missing data may
contribute to the elevated frequency of homoplasic charac-
ters. We are thus of the opinion that adequate sampling of
specimens may further increase the performance of the
NORs in resolving phylogenetic relationships in Bovidae,
possibly also at higher taxonomic levels.
The availability of dated divergences within Bovidae
(Hassanin and Douzery, 1999a, b) allowed us to place the
NOR data in a temporal framework and to highlight why
these data are phylogenetically useful in some taxonomic
groups, and not in others. In our opinion the data on the
chromosomal distribution of NORs can be valuable phylo-
genetic characters when at least three conditions are met.
First, their number per karyotype should not be very low,
and their chromosomal location should vary among spe-
cies. Secondly, NOR variability should show a diversity of
evolutionary ages and, fina lly, the group of taxa under study
should be characterized by a low to moderate rate of karyo-
type reorganization. When these conditions are met, NORs
become phylogenetically useful characters comparable to
structural rearrangements.
Chromosomal vs. molecular phylogeny
Molecular phylogenetic studies involving concatena-
tions of mitochondrial and/or nuclear genes all provide a
strong support for monophyly of the subfamily Bovinae and
the recognition of three tribes, the Bovini (cattle and buf-
faloes), Tragelaphini (African spiralled-horned bovids) and
Boselaphini (nilgai and chousingha) (Hassanin and Dou-
zery, 1999a, b; Hassanin and Ropiquet, 2004 and references
therein). However, the intertribal relationships remain
poorly resolved with the most recent analysis suggesting
two alternative hypotheses for these associations, one a sis-
ter-group relationship between Bovini and Tragelaphini,
and another one favoring an association of Boselaphini and
Tragelaphini (Hassanin and Ropiquet, 2004).
In terms of the saola’s phylogenetic affinities, all molecu-
lar studies agree on its placement within the subfamily Bo-
vinae (Dung et al., 1993; Hassanin and Douzery, 1999b;
Gatesy and Arctander, 2000; Hassanin and Ropiquet, 2004).
There is no consensus, however, in terms of its tribal affini-
ties with Dung et al. (1993) suggesting a placement in Bose-
laphini, whereas Hassanin and Douzery (1999b), Gatesy
and Arctander (2000) and Hassanin and Ropiquet (2004)
find an evolutionary affinity with Bovini. The latter authors
define three subtribes within Bovini, (i) Bovina which in-
cludes all species of Bos and Bison , (ii) Bubalina which
groups Bubalus and Syncerus, and (iii) Pseudoryina, which
is represented only by the saola, P. nghetinhensis . The inter-
relationships among the Bovini subtribes remain uncertain
and only tentatively saola was considered as sharing closer
phylogenetic affinities with the subtribe Bovina (Hassanin
and Douzery, 1999b; Gatesy and Arctander, 2000; Hassanin
and Ropiquet, 2004). The topological conflicts outlined
above are summarised on the consensus phylogenetic tree
( Fig. 9 b).
Despite a rather limited number of phylogenetically in-
formative chromosomal characters identified and relatively
weak support of some groupings on the chromosomal tree
( Fig. 9 a), the cytogenetic data nevertheless allow an evolu-
tionary scenario which, although largely congruent with
that inferred from molecular analyses ( Fig. 9 b), differs in
certain important aspects allowing an interpretation that
suggests the following sequence of events. After the Anti-
lopinae/Bovinae split, three main lineages emerged within
the subfamily Bovinae: one leading to Boselaphus , another
one to Tragelaphini, and the last to Bovini; the three tribes
differ from each other by their particular X:autosome con-
figurations. The ancient type of the bovine X ( Fig. 6 , II) was
modified by a gonosome-autosome translocation involving
BTA14 ( Fig. 6 , V) in Boselaphus (unfortunately the sex chro-
mosome constitution of chousingha Tetracerus quadricor-
nis remains unstudied), and BTA13 in the species of the
tribe Tragelaphini ( Fig. 6 , IV). Since the establishment of
these potentially negatively heterotic rearrangements, chro-
mosomal evolution in these lineages occurred indepen-
dently. The chromosomal data suggest that the first lineag-
es to diverge were Boselaphus and the Tragelaphini since
both have NORs on BTA5 (character 49) which is shared
with the more distant Caprini, but which is lost in all other
in-group taxa. This divergence also predated the emergence
of the tribe Bovini with its two monophyletic clades, one
corresponding to the subtribe Bovina sensu Hassanin and
Ropiquet (2004) which is strongly supported by modifica-
tion of the X chromosome ( Fig. 6 , III) and the emergence of
the new NORs site (character 53), and another uniting spe-
cies of the subtribe Bubalina sensu Hassanin and Ropiquet
(2004) plus the saola, P. nghetinhe nsis . The last grouping is
supported by emergence of a new NOR site on BTA22 (char-
acter 58), and the loss of the old one on BTA2 (char. 46).
Although interspecies relationships within the subtribe Bo-
vina are unresolved in our analysis, they are clearly defined
within Bubalina where S. caffer was the first species to di-
verge, an event that is underscored by a series of Rb fusions
(characters 13 to 16) that are likely to have resulted in an ef-
fective genetic isolation. The remaining species, Bubalus
bubalis , B. depressicornis , B. mindorensis and P. ng het inhen -
sis , share a newly acquired NOR site on BTA26 (character
61), but were subsequently differentiated by a series of 5 Rb
translocations that are specific to the saola (characters 22 to
26), two (characters 2 and 5) that are unique to two of the
three species of the genus Bubalus studied, and two charac-
Cytogenet Genome Res 122:41–54 (2008) 53
ters (6 and 7) that are shared by all three genera. Addition-
ally, the saola could be definitely excluded from the subtribe
Bovina on the morphology of its X chromosome. In this
context, the three NOR characters shared by saola and spe-
cies of the subtribe Bubalina, i.e., 58 (all but Bubalus depres-
sicornis ), 59 (with Bubalus depressicornis ) and 61 (all but
Syncerus caffer ), strongly suggest its placement within this
subtribe, and the recognition of two rather than three sub-
tribes within the tribe Bovini.
In conclusion, our study highlighted the usefulness of
NORs for phylogenetic reconstructions within Bovidae,
while at the same time emphasizing the importance of ap-
propriate sample sizes to rule out variation in their numbers
and chromosomal location. We outline the conditions un-
der which these chromosomal structures may result in ad-
ditional phylogenetic information. Finally our data suggest
the importance of X chromosome rearrangements in evolu-
tion of the species studied herein and raise the possibility
that detailed molecular cytogenetic analysis of the sex chro-
mosomes may provide additional insights on evolution and
phylogenetic relationships within Bovidae and probably
Ruminantia.
A c k n o w l e d g m e n t s
The authors thank N.T. Uoc for the saola f ibroblast cells and D.
Tuoc, H.N. Khanh, H.V. Keo for providing help to collect the saola
samples. We are grateful to B. Dutrillaux and F. Richard for providing
facilities and help for the study. Our special tha nks go to T.J. Robinson
for very useful comments on the earlier draft of the manuscript and its
linguistic improvement.
Appendix
Chromosomal characters identif ied in Bovidae. Rob = Robertsonian translocation, t = tandem fusion, inv = inversion, t rcp = reciprocal
translocation, NOR = nucleolus-organizing region bearing chromosomes. Chromosomes of studied species were numbered according to cattle
standard nomenclature (ISCNDB2000).
1. rob (2;28); 2. rob (5;28); 3. t (7;5;28); 4. t (11;5;28); 5. rob (1;27); 6. rob (2;23); 7. rob (8;19); 8. rob (16;25); 9. rob (11;20); 10. rob (17;25);
11. rob (4;14); 12. rob (16;29); 13. rob (1;13); 14. rob (2;3); 15. rob (5;20); 16. rob (11;29); 17. rob (1;5); 18. rob (6;13); 19. rob (8;12); 20. rob (19;27);
21. rob (24;25); 22. rob (1;10); 23. rob (8;13); 24. rob (6;19); 25. rob (4;18); 26. rob (11;12); 27. inv para12; 28. t(3; 22;2); 29. rob (5;10); 30. rob (6;11);
31. rob (1;25); 32. rob (4;12); 33. rob (8;24); 34. rob (9;20); 35. rob (7;27); 36. rob (15;16); 37. rob (18;19); 38. rob (14;26); 39. rob (21;23); 40. rob (17;28);
41. t rcp [(9;20);(14;26)]; 42. rob (1;3); 43. rob (2;8); 44. rob (5;11); 45. t rcp (9;14); 46. NOR BTA2; 47. NOR BTA3; 48. NOR BTA4; 49. NOR BTA5;
50. NOR BTA7; 51. NOR BTA8; 52. NOR BTA10; 53. NOR BTA11; 54. NOR BTA14; 55. NOR BTA16; 56. NOR BTA18; 57. NOR BTA19; 58. NOR
BTA22; 59. NOR BTA23; 60. NOR BTA25; 61. NOR BTA26; 62. NOR BTA28; 63. inv (X); 64. transposition (X); 65. t (Y;13); 66. t (X;14), t (Y;14)
Matrix of chromosomal characters
111111111122222222223333333333444444444455555555556666666
123456789012345678901234567890123456789012345678901234567890123456
BTA 000000000000000000000000000000000000000000000111000010000001011100
BFR 100000000000000000000000000000000000000000000111000010000001001100
BJA 000000000000000000000000000000000000000000000111000010000001001100
BIN 000000000000000000000000000000000000000000000111000010000001001100
BBO 000000000000000000000000000000000000000000000110000010000001011100
BBI 000000000000000000000000000000000000000000000111000010000001011100
SCA 000000000000111100000000000000000000000000000010000000000101011000
BBU 011011110000000000000000000000000000000000000010000000001101111000
BDE 010011101100000000000000000000000000000000000010000001001011101000
BMI 000101100011000000000000000000100000000000000010000000001100111000
PNG 000000000000000000000111111000000000000000000000011001010110101000
BTR 000000000000010011111000000000000000000000000100100001000000001001
TOR 000000000000000000000000000111111111111100000110100100100000011010
TDE 000000000000000000000000000111111011101110000?????????????????1010
RRU 000000000000000000000000000000000000000001001111100000000000010000
CIB 000000000000000000000000000000000000000000001111100000000000010000
OAR 000000000000000000000000000000000000000001111111100000000000010000
References
Ahrens E, Graphodatskaya D, Nguyen BX, Stran-
zinger G: Cy togenetic comparison of saola
( Pseudoryx nghetinhensis ) and cattle ( Bos tau-
rus ) using G- and Q-banding and FISH. Cy to-
genet Genome Res 111:
147–151 (2005).
Bloom SE, Goodpa sture C: An improved t echnique
for selective silver staining of nucleolar orga-
nizer regions in huma n chromosomes. Hum
Genet 34:
199–206 (1976).
Bosma AA, de Ha an NA, Arkest eijn GJ, Yang F, Yer-
le M, Zijlstra C: Comparative chromosome
painting between the domestic pig ( Sus scrofa )
and two sp ecies of pecca ry, the colla red peccar y
( Ta ya s su ta jac u ) and the white-lipped peccary
( T. pe ca ri ): a phylogenetic perspective. Cyto-
genet Genome Res 105:
115–121 (2004).
Cytogenet Genome Res 122:41–54 (2008)54
Buckland RA, Evans HJ: Cytogenetic aspects of
phylogeny in the Bovidae. I. G -banding. Cyto-
genet Cell Genet 21:
42–63 (1978).
Caspersson T, De la Chapelle A, Schroeder J, Zech
L: Quinacrine f luorescence of metaphase ch ro-
mosomes. Identical patterns in different tis-
sues. Ex p Cell Res 72:
56–59 (1972).
De la Seña C, Chowdhary BP, Gustavsson I: Local-
ization of t he telomeric (TTAGGG)n sequenc es
in chromosomes of some domestic animals by
fluorescence in situ hybridization. Hereditas
123:
269–274 (1995).
Dobig ny G, Ozouf-Costa z C, Bonillo C, Volobouev
V: ‘Ag-NORs’ are not always true NORs: new
evidence in mammals. Cytogenet Genome Res
98:
75–77 (2002).
Do bigny G, Granjon L, A niskin V, Bâ K, Volobouev
V: A new sibling spec ies of Taterillus (Rodentia,
Gerbil linae) from West Africa. Mam m Biol 68:
299–316 (2003a).
Dobig ny G, Ozouf-Costa z C, Bonillo C, Volobouev
V: Evolution of rRNA gene clusters and telo-
meric repeats during explosive genome repat-
terning in Tat eri llu s X (Rodentia, Gerbillinae).
Cytogenet Genome Res 103:
94–103 (2003b).
Dobigny G, Ducroz JF, Robinson TJ, Volobouev V:
Cytogenetics and cladistics. Syst Biol 53:
470–
484 (2004).
Dung V V, Giao PM, Chinh NN, Tuoc D, Mac Kin-
non J: A new species of living bovid from Viet-
nam. Nature 363:
443–445 (1993).
Gallag her DS, Womack JE: Chromosome conser-
vation in t he Bovidae. J Hered 83:
287–298
(1992).
G allagher DS , Davis SK, De Donato M, Bur zlaff JD,
Womack JE, et al: A karyotypic a nalysis of nil-
gai, Boselaphus tragocamelus (Artiodact yla:
Bovidae). Chromosome Res 6:
505–513 (1998).
G allagher DS , Davis SK, De Donato M, Bur zlaff JD,
Womack JE, et al: A molecula r cytogenetic
analysis of the tribe Bovini (Artiodactyla: Bo-
vidae: Bovinae) with an emphasis on sex chro-
mosome morphology and NOR dist ribution.
Chromosome Res 7:
481–492 (1999).
Gatesy J, Arctander P: Hidden morphological sup-
port for the phylogenetic placement of Pseudo-
ryx nghetinhensis with bovine bovids: a com-
bined ana lysis of gross anatomica l evidence
and DNA sequences from five genes. Syst Biol
49:
515–538 (2000).
Gerbault-Serreau M, Bonnet-Ga rnier A, Richard F,
Dutrillaux B: Chromosome painting compari-
son of Leontopithecus chrysomelas (Ca ll itrich-
ine, Platyrrhi ni) with man and its phylogenetic
position. Chromosome Res 12:
691–701 (2004).
Hassanin A, Douzer y EJP: The tribal rad iation of
the family Bovidae (Art iodact yla) and the evo-
lution of the m itochondrial c ytochrome b gene.
Mol Phylogenet Evol 13:
227–243 (1999a).
Hassanin A, Douzery EJP: Evolutionary affinities
of the enigmatic saola ( Pseudoryx nghetinhen-
sis ) in t he context of the molec ular phylogeny of
Bovidae. Proc R Soc Lond B 266: 893–900
(1999b).
Hassanin A, Douzery EJ P: Molecula r and morpho-
logical phylogenies of Ruminantia and the a l-
ternative position of the Moschidae. Syst Biol
52:
206–228 (20 03).
Hassanin A, Ropiquet A: Molecular phylogeny of
the tr ibe Bovini (B ovidae, Bov inae) and the ta x-
onomic status of the Kouprey, Bos sauveli Ur-
bain 1937. Mol Phylogenet Evol 33:
896–907
(20 04).
IS CNDB (2000): Internation al System for Chromo-
some Nomenclatu re of Domestic Bovid s. Di Be-
rardino D, Di Meo GP, Gallagher DS, Hayes H,
Iannuz zi L (eds). Cytogenet C ell Genet 92:
283–
299 (2001).
IUCN (International Union for Conservation of
Nature and Nat ural Resources) IUCN Red List
of Threatened Species. w ww.iucnredlist.org
(2007).
Ki ng M: Species Evolution: the Role of Chromo-
some Change ( Cambridge University Press,
Cambridge, UK 1993).
Kumamoto AT, Charter SJ, Houck ML, Fra hm M:
Chromosomes of Damaliscus (Artiodac tyla,
Bovidae): simple a nd complex centric f usion re-
arrangements. Chromosome Res 4:
614–621
(19 96).
Mallon DP, Kingswood SC: Antelopes – Global Sur-
vey and Regional Action Plans, Part 4: North
Afri ca, the Midd le East, and A sia (IUCN, Gland
2001).
Mayr B, Tesarik E, Auer H, Burger H: Nucleolus-
organizer regions and heterochromatin in
three species of Bovidae. Genetica 75:
207–212
(198 7).
Mey ne J, Baker RJ, Hobar t HH, Hsu TC, Ryder OA,
et al: Distribution of non-telomeric sites of t he
(TTAGGG)n telomeric sequence in vertebrate
chromosomes. Chromosoma 99:
3–10 (1990).
Ng uyen TT, Nguyen BX, Stran zinger G: Chara cter-
ization of G-banded chromosomes of a female
Saola ( Pseud oryx ngheti nhensis ) and X chromo-
some identification by means of fluorescent in
situ hybridi zation (FISH). Cy togenet Genome
Res 109:
195–198 (2005).
O’Brien SJ, Menninger JC, Nash WG (eds): Atlas
of Mamma lian Chromosomes (John Wiley &
Sons, Hoboken 2006).
Pinton A, Ducos A, Séguéla A, Berland HM, Darré
R, Dar ré A, et al: Chara cterization of re ciprocal
translocations in pigs using dua l-colour chro-
mosome painting and primed in sit u DNA la-
belling. Chromosome Res 6:
361–366 (1998).
Ponce de Leon FA, Ambady S, Hawk ins GA, Kappe s
SM, Bishop MD, et a l: Development of a bovine
X chromosome linkage group and painting
probes to assess cattle, sheep, and goat X chro-
mosome segment homologies. Proc Nat l Acad
Sci USA 93:
3450 –3454 (1996).
Robichaud WG: Physical and behavioral descrip-
ti on of a ca pt iv e s aol a, Pse udoryx nghe tinhensis .
J Mamma l 79:
394–4 05 (1998).
Robinson TJ, Wilson V, Gallagher DS, Taylor JF,
Davis SK, e t al: Chromosoma l evolution in dui-
ker antelope (Cephalophinae: Bovidae): karyo-
type comparisons, fluorescence in situ hybrid-
ization, a nd rampant X chromosome v ariation.
Cytogenet Cell Genet 73:
116–122 (19 96).
Ro binson TJ, Harr ison WR, Ponce de Lé on A, Elder
FFB: X chromosome evolution in t he suni and
eland antelope: detection of homologous re-
gions by f luorescence in situ hybridization and
G-banding. Cytogenet Cell Genet 77:
218–222
(19 97).
Robi nson TJ, Harris on WR, Ponce de Leon FA, Da-
vis SK, Elder FFB: A molecular cy togenetic
analysis of X chromosome repatterning in the
Bovidae : transposition s, inversions, and phylo -
genetic inference. Cytogenet Cell Genet 80:
179–18 4 (1998).
Ropiquet A, Hassanin A: Molecular phylogeny of
Caprines (Bovidae, A ntilopinae): the question
of their origin and d iversification during the
Miocene. J Z ool Syst Evol Res 43:
49–60 (2005).
Rousseaux S, Chev ret E, Montei l M, Cozzi J, Pel le-
tier R, et a l: Meiotic segregation in males het-
erozygote for reciprocal translocations: ana ly-
sis of sperm nuclei by two and t hree colour
fluorescence in situ hybridization. Cytogenet
Cell Genet 71:
240–246 (1995).
Sc haller G, Rabi nowitz A: The saola or spi ndlehorn
bovid Pseudoryx nghetinhensis in Laos. Oryx
29:
107–114 (1994).
S eabright M: A rapid band ing technique for hu man
chromosomes. Lancet 2:
971–972 (1971).
Searle JB: Chromosomal hybrid zones in eutherian
mamma ls, in Ha rrison RG (ed): Hybrid Zones
and the Evolutionary Process (Oxford Univer-
sity Press, Oxford 1993).
Sumner AT: A simple technique for demonst rating
centromeric heterochromatin. Exp Cell Res 75:
304–306 (1972).
Sumner AT: Chromosome Banding (Unwi n Hy-
man, London 1990).
Sumner AT: Chromosomes: Organization and
Function (Blackwell Publishing Ltd, Oxford
2003).
Swofford DL: PAUP * Phylogenetic A nalysis Using
Parsimony a nd Other Method, version 3.1
(Sinauer Associates, Sunderland 1993).
Tanak a K, Solis CD, Ma sangkay JS , Maeda K, Kawa-
moto Y, Namikawa T: Phylogenetic relation-
ship among a ll livi ng species of the genu s Buba-
lus b ased on DNA sequences of t he cytochr ome
b gene. Biochem Genet 34:
443– 452 (1996).
Tanaka K, Matsuda Y, Masang kay JS, Solis CD,
Anuncia do RV, et al: Cyto genetic ana lysis of the
tamaraw ( Bubalus mindorensis ): a comparison
of R-banded karyoty pe and chromosomal dis-
tribution of centromeric satellite DNAs, telo-
meric sequence, and 18S-28S rR NA genes with
domestic water buffa loes. J Hered 91:
117–121
(2000).
Taylor PJ: Patte rns of chromosomal v ariation in A f-
rican rodents. J Mammal 81:
317–331 (2000).
T homas H: Crania l anatomy and phylogenet ic rela-
tionships of a new bovine ( Pseudoryx nghetin-
hensis ) d iscovered in the Vietnamese Anna-
mite mountain range. Mammalia 58:
453–481
(19 94).
Vassa rt M, Séquéla A, Haye s H: Chromosomal evo -
lution in gazelles. J Hered 86:
216–227 (1995).
Volleth M: Differences in the location of nucleolus
organizer regions in European vespertilionid
bats. Cytogenet Cell Genet 44:
186–197 (1987).
Volobouev V, Aniskin V, Lecompte E, Ducroz JF:
Pattern s of karyoty pe evolution in c omplexes of
sibling species within three genera of African
murid rode nts inferred from t he comparison of
cyto genetic and molecula r data. Cytogene t Ge-
nome Res 96:
261–275 (2002a).
Volobouev V, Ducroz JF, Aniskin V, Britton-David-
ian J, Castiglia R, et al: Chromosomal charac-
terisation of Arvicanthis species (Rodentia:
Murinae) from Western and Central Africa:
implications on taxonomy. Cytogenet Genome
Res 96:
250–260 (2002b).
Volobouev V, Aniskin V, Sicard B, Dobigny G,
Granjon L: Systematics and phylogeny of West
African gerbi ls of the genus Gerbilliscus (Muri-
da e: G erbi ll ina e) in ferr ed f rom c ompa rat ive G-
and C-banding chromosomal analyses. Cyto-
genet Genome Res 116:
269–281 (2007).
Wurster DH, Benirschke K: Chromosome studies
in the superfami ly Bovoidea. Chromosoma 25:
152–171 (1968).
Zijlstra C, Mellin k CH, de Haan NA, Bosma A A:
Localization of the 18S, 5.8S and 28S rRNA
genes and the 5S rRNA genes in t he babirusa
and the white-lipped peccary. Cytogenet Cell
Genet 77:
273–277 (1997).
