Gorilla mtDNA - Sequences Unravelled and Secrets Revealed
Non-invasive genetics, the analysis of DNA variability using samples
such as hair or fresh faeces collected in the field, has made enormous
progress since the days of its infancy in the late 1980s and early 1990s
(Higuchi et al. 1988; Constable et al. 1995). At this time, such genetic
analyses held great promise for ecologists, with the capacity to explain
all the mysteries of population structure and social organisation, degrees
of relatedness and gene flow of species not amenable to direct study (Avise
1994). We were no less enthusiastic in our predictions of its power and
our prowess. So much so that we convinced a pool of gorilla researchers
and conservation organisations to help us collect material for the first
range-wide study of genetic variability in the gorilla.
Traditionally three subspecies have been recognised within a single species,
Gorilla gorilla (Groves 1967, 1970): western lowland (G. g.
gorilla), eastern lowland gorillas (G. g. graueri) and mountain
gorillas (G. g. beringei). A more recent evaluation of the available
data has led to a reclassification of the gorilla into two species (Groves
2001): the western gorilla Gorilla gorilla and the eastern gorilla
Gorilla beringei. Within western gorillas, two subspecies have
been proposed: G. g. gorilla (western gorillas except those in
the Cross River area between Nigeria and Cameroon) and G. g. diehli
(Cross River gorillas). Within eastern gorillas, three subspecies have
also been proposed: G. b. graueri (eastern lowland), G. b. beringei
(Virunga mountain gorilla) and, perhaps, an as yet unnamed third taxonomic
unit from the Bwindi forest, Uganda.
Nearly 10 years later we believe we have achieved our goal (Clifford et
al. 2004), and with this article I would like to acknowledge all the people
who have been instrumental in this achievement (see list of field collaborators).
In fact our aims were three-fold: in parallel to the question of gorilla
population genetics we were committed to developing a regional molecular
ecology laboratory where African scientists could receive training and
to providing a facility where samples could be treated without having
to leave the region. The Centre International de Recherches Médicales
de Franceville (CIRMF) still remains the only centre where such work can
be carried out. Too many studies in the Central African region have involved
the collection of material, its export to northern institutions and analyses
carried out to satisfy the requirements of a thesis, without reference
to the development of local expertise. Equally, having realised that it
is not feasible to attempt automated genetic analyses in Gabon without
the difficulties associated with equipment maintenance and delivery, we
have developed a strategy, in collaboration with overseas laboratories,
whereby fundamental molecular work is carried out at CIRMF, leaving costlier
and more technologically challenging work to collaboration.
Through such collaboration, local scientists and students receive technical
training and scientific support for their research both in-country and
overseas. For example, sample extraction, species and sex determination,
and mitochondrial DNA (mtDNA) screening are carried out in the UGENET
laboratories at CIRMF, and DNA for automated sequencing and microsatellite
techniques is analysed through collaborating laboratories. Such high cost,
high throughput, technically challenging work on non-invasive material
is best done through such collaborative networks. By this means, all samples
can remain in the region providing valuable resources for further study
without compromising future capacity for host country research and development.
Biological material collected non-invasively provides the only means of
gleaning information on feeding ecology, habitat preference and now genetic
population structure for a number of cryptic species living in habitat
conditions unsuitable for direct study. Gorillas conveniently leave behind
night nests and faecal deposits, which have, in other species, been used
as a source of DNA for genetic work (Morin et al. 1994). Hairs collected
from night nests were initially thought to be an ideal source of DNA,
but forensic based work by Kathryn Jeffery (2003) in Cardiff, UK, has
now shown that many shed hairs are in fact largely devoid of cellular
material from which DNA could be extracted quantitatively. Molecular techniques,
however, have advanced so far as to be able to amplify the tiny (pg) amounts
of mitochondrial and genomic DNA associated with these hairs and faeces
(Morin et al. 2001).
We now know, however, that these very small amounts of DNA, degraded into
short fragments on exposure to humid forest conditions, can lead to genotyping
errors from allelic dropout (stochastic non-amplification of one allele
over the other; Taberlet et al. 1996), and false amplification of non-allelic
artefacts. Mitochondrial DNA (mtDNA) analyses are further complicated
by the translocation of mtDNA fragments into the nuclear genome (Numts),
where they undergo a separate evolutionary history and can consequently
confound the analysis of true mitochondrial phylogenies (Collura &
Stewart 1995). Other errors in sequence interpretation can arise through
in vitro recombination between mitochondrial and nuclear translocated
fragments (Thalmann et al. 2004) and heteroplasmic mutations that appear
common in the hyper-variable control region (e.g. Tully et al. 2000).
The presence of a poly-cytosine (poly-C) rich tract within the first hyper-variable
region (HV-1) has also proved difficult to align and sequence, further
complicating the analysis of this region. Recently, one research group
has stated that, due to all these potential errors, it is impossible to
place any confidence in studies using mtDNA to characterise genetic variability
within gorillas (Thalmann et al. 2004; Vigilant et al. 2004).
So where does this leave us? What future is there for molecular ecologists
without access to biological material yielding high quality DNA where
longer sequences of DNA can be generated and authenticated? Or for studies
of recent evolution of populations, based on sex-specific gene flow if
maternally inherited mtDNA is deemed unreliable? Must we abandon all hope
of studying the phylogeography of species for which we have no alternative
but to use low quality DNA from non-invasive samples? Could we not put
this knowledge of mtDNA variation present within one and the same individual
to some use?
MtDNA sequence variation provides a powerful means of understanding genetic
variation and evolution, and mtDNA sequences have been the primary source
of data for resolving questions about modern human origins (Ruvolo et
al. 1994) and the sub-specific genetic variation of extant chimpanzees
(Gagneux et al. 2001) and gorillas (Garner & Ryder 1996; Jensen-Seaman
& Kidd 2001). In particular, analyses have focused on the control
region that is involved in the control and initiation of replication.
This region has an extremely high rate of mutation, and in most species
has characteristic motifs that allow sequences to be aligned into specific
haplogroups.
Rather than class nuclear translocation of mitochondrial genes as confounding
factors in the interpretation of mtDNA variation, we have tried to profit
from their existence, by recognition of the fact that they can be characterised
into different groups, which can then be used as evolutionary markers
in their own right. The methods we used for identifying Numts were as
exhaustive as we could afford to make them and relied heavily on cloning
of PCR products with sequencing of multiple clones from one individual.
A combination of phylogenetic analysis, poly-C domain sequence motifs
and diagnostic sites in the region flanking this domain allowed us to
(i) discriminate between putative Numt DNA sequences and their presumed
mtDNA counterparts, and (ii) classify Numts into different categories.
We could provisionally identify Numts not only in the data set we generated
ourselves, but also in sequences deposited in GenBank, the central repository
for genetic sequence data, that were previously classified as authentic
mtDNA.
From the several hundred hair samples received from 20 different sites
of the gorilla range we were able to generate 53 complete sequences of
258 bp of the HV-1 region of the control region from gorillas throughout
the eastern and western ranges. An additional 30 sequences from 3 new
and 5 sampled sites were retrieved from GenBank. Of these 83 sequences,
59 were deemed to be true mtDNA sequences; 16 from eastern gorillas and
43 from western gorillas. The remaining 24 sequences (14 from GenBank)
were classified as Numts. Work in progress is now examining an even larger
data set using additional sequence data derived from sites in our initial
study and additional sites throughout Gabon.
 |
|
Geographic distribution and haplogroup designation (A-D) of sequences sampled from 23 sites across current gorilla range. The area of each circle is proportional to number of sequences analysed at each site and proportionally divided where more than one haplogroup is present. The present day geographical distribution of gorillas is shaded in grey. Subgroups are reflected in circle coloration. Reproduced by permission of Blackwells Publishing
from Clifford et al. (2004) |
 |
Minimum spanning network of pair-wise absolute differences between gorilla mitochondrial DNA haplotypes. For three-letter codes see Figure 1. The area of each circle represents the pro-portional representation of each of the respective haplotypes. Branch lengths are also proportionally represented and hash marks for closely related haplotypes indicate individual mutational steps. Haplogroups A to D are colour coded and subgroups C1, C2, D1, D2 and D3 are indicated. Reproduced by permission of Blackwells Publishing from Clifford et al. (2004) |
|
| Bwindi: BWD, Kahuzi-Biega: KBG, Itombwe: ITW, Tshiaberimu: TSH, Lobéké: LBK, Equatorial Guinea: EQG, Central African Republic/Lobéké/Ndoki: CAR/LBK3/NDK1 and Gabon/Congo: GAB/CON, Belinga: BEL, Conkouati: CQT, Itombe: ITO, Lopé: LOP, Lastourville: LAS, Lossi: LOS, Petit Loango: PLO, Rabi: RAB. Haplogroups A to D and subgroups C1, C2, D1, D2 and D3 are indicated. Reproduced by permission of Blackwells Publishing from Clifford et al (2004) | ||
Having identified and subsequently excluded Numt sequences from the
analysis, 4 major mtDNA haplogroups were identified (A-D), comprising
a total of 36 unique mitochondrial haplotypes at 23 different sites of
the gorilla range. Haplogroups A and B correspond to mountain and eastern
lowland gorillas, respectively. Haplogroups C and D together cover the
western lowland gorilla range, with C spanning from the Cross River region
in Nigeria/Cameroon, through Dja and Lobéké in Cameroon
to Ipassa in northeastern Gabon and one museum sample from the Uele Valley,
Democratic Republic of the Congo. Haplogroup D encompasses Gabon, Congo,
Central African Republic and Equatorial Guinea, as well as one museum
sample from Cameroon. The separation between eastern (A, B) and western
(C, D) haplogroups recapitulates the large genetic distance and major
evolutionary split between eastern and western gorillas, with mountain
gorillas (A) distinct from eastern lowland gorilla populations (B) (Garner
& Ryder 1996).
The most striking finding in this study is the identification of two distinct
groups within western lowland gorillas (haplogroups C and D). Genetic
divergence in the mitochondrial control region between the two western
groups C and D is on average greater than that seen between the two eastern
haplogroups A and B, which are presented in the new taxonomic description
as two different subspecies (Groves 2001). This divergence within western
lowland gorillas does not coincide with any previously recognised biogeographic
barrier, but may potentially be linked to historic climatic events and
changes in forest cover (see later). In addition, western gorillas are
more genetically diverse within each haplogroup than are the two eastern
groups, and sub-structuring is evident within both groups C and D. The
Cross River gorillas (G. g. diehli) belong to the most diverse
haplogroup (C), which also includes gorillas from south of the Sanaga
River in southern Cameroon and from adjacent northeastern Gabon.
Within the two western lowland gorilla haplogroups, two major subgroups
are evident in haplogroup C (C1, C2) and 3 geographically partitioned
subgroups in haplogroup D (D1, D2, D3). Subgroup D1 comprises gorillas
almost exclusively from Equatorial Guinea, subgroup D2 constitutes gorillas
from Central African Republic, northern Congo and one sample from Lobéké,
and subgroup D3 comprises the majority of gorillas from Gabon and adjacent
Congo. Interestingly, subgroup D3, which covers the largest surface area
and contains the largest number of gorillas, also shows the lowest genetic
diversity of the 5 subgroups.
Patterns of genetic variation indicate that a history of population fragmentation
may have given rise to the distinct haplogroups identified in this study.
Mismatch distributions provide limited evidence of demographic expansion
in eastern lowland gorilla populations as observed in a previous study
(Jensen-Seaman & Kidd 2001).
Within western gorillas, sub-groups D2 (CAR) and D3 (Gabon/adjacent Congo)
show evidence of expansion whereas gorilla populations in Nigeria/Cameroon
(C) exhibit a more complex population structure and history. Periodic
changes in climate during recent Pleistocene history led to repeated retractions
of vegetation cover into isolated refugia during glacial maxima (Maley
1996). Distributions of species dependent on closed canopy forest would
have followed these changes, leading to population fragmentation within
restricted forest refugia, from which expansion would later follow during
climate warming. Such repeated isolation and expansion events may have
had profound effects on genetic structure, as demonstrated in western
gorillas. Several montane refuges have been identified in western Central
Africa (Maley 1996), and the existence of fluvial refuges has also been
proposed (Colyn 1991). The present location of D2 would correspond to
one of these fluvial refuges, whereas the remaining subgroups could be
traced to montane forest remnants in Cameroon and Gabon/Equatorial Guinea.
Riverine barriers do not appear to have influenced gorilla history to
the same extent.
Recognised barriers such as the Sanaga River (Grubb 2001) have had no
apparent effect on gorilla divergence; rather the occurrence of haplogroups
C and D in Lobéké (Cameroon) may reflect recent gene flow
between adjacent haplogroups across the Sangha River following recent
post-glacial expansion. Similarly admixture consistent with ongoing population
expansion out of Nigeria/Cameroon, and refuges in Gabon (Monts de Cristal
and Massif du Chaillu), could explain the diversity of types found in
northeastern Gabon.
What are the implications of these findings for gorilla conservation?
The relatively deep subdivision between haplogroups C and D within western
gorillas in conjunction with the divisions between eastern and mountain
gorillas would support the recognition of four distinct evolutionary significant
units (ESU, Moritz 1994). It might be premature to base such conclusions
on mtDNA diversity alone, given the stochasticity of a single marker system
and the fact that neutral markers may fail to detect divergence in ecologically
important traits. Nevertheless, "demes" based on morphological
traits identified in western gorillas (Groves 1967, 1970) correspond by
and large to the geographical separation seen in genetic signatures, although
the morphologically distinct Cross River gorillas in Nigeria belong to
a larger haplogroup encompassing gorillas in Cameroon. The conservation
status of all gorillas within group C appears equally precarious, due
to extreme habitat fragmentation and human pressure (Oates 2002; Groves
2002).
This study demonstrates that authentic mitochondrial genetic diversity
can be assessed in the context of biological and analytical artefacts
such as Numts, heteroplasmy and in vitro recombination, and future work
will clarify the importance of heteroplasmy and nuclear integrations as
evolutionary markers in their own right. In a historical and biogeographical
context, our results show that distribution of forest cover during the
recent past may have had profound effects on the divergence of gorilla
populations, and we would suggest that conservation policy should aim
to preserve these regional differences.
E. Jean Wickings, Stephen L. Clifford, Nicola M. Anthony, Kathryn Jeffery, Mireille Johnson-Bawe, Katherine A. Abernethy and Michael W. Bruford
Field collaborators: Kahuzi-Biega, D. R. Congo: D. Bonny, K. P. Kiswele (CRSN), I. Omari, C. Sikubwabo (ICCN), L. White, J. Hall, I. Bila-Isia, H. Simons Morland, E. Williamson, K. Saltonstall, A. Vedder, K. Freeman, B. Curran (WCS) J. Yamagiwa (Kyoto Univ.); Itombwe, D. R. Congo: I. Omari, F. Bengana (ICCN); J. Hart (WCS); Concouati, Congo: B. Goossens (Univ. of Cardiff), A. Jamart (HELP); Rabi, Gabon: S. Lahm (IRET); Petit Loango, Gabon: J. Yamagiwa (Kyoto Univ.); Lopé, Gabon: C. Tutin, K. Abernethy, E. Dimoto, J. T. Dinkagadissi, R. Parnell, P. Peignot, B. Fontaine (CIRMF), M. E. Rogers, L. White, B. Voysey, K. McDonald, (Edinburgh), R. Ham (Stirling), J. G. Emptaz-Collomb; Lastourville, Gabon: Y. Mihindou (WCS-MIKE); Ipassa and Belinga, Gabon: S. Lahm, J. Okouyi (IRET); Itombe, Gabon: P. Telfer (NYU); Lossi, Congo: M. Bermejo, G. Illera, F. Maisels (ECOFAC); Bai Hokou, Central African Republic: M. Goldsmith (Tufts Univ.), L. White (WCS); Nouabalé-Ndoki, Congo: P. Walsh (WCS); Lobéké, Cameroon: L. White, L. Usongo (WCS); Dja, Cameroon: E. Williamson (ECOFAC), L. Usongo (WCS/ECOFAC); Afi Mts./Cross River, Nigeria: K. McFarland, J. Oates (CUNY, USA). E. Nwufoh (CRNP); Monte Alen, Equatorial Guinea: M. Bermejo, G. Illera (ECOFAC); Belar, Cameroon: M. Harman (Powell-Cotton Museum)
References
Avise J. C. (1994) Molecular Markers, Natural History and Evolution. New
York (Chapman and Hall)
Clifford, S. L. et al. (2004) Mitochondrial DNA phylogeography of western
lowland gorillas (G. g. gorilla). Mol. Ecol. 13: 1551-1565
Collura, R. V. & Stewart, C.-B. (1995) Insertions and duplications
of mtDNA in the nuclear genomes of Old World monkeys and hominoids. Nature
378: 485-489
Colyn, M. (1991) L'importance Zoogéographic du Basin du fleuve
Zaïre pour la spéciation. Annales Sciences Zoologiques 264:
180-185
Constable, J. J. et al. (1995) Nuclear DNA from primate dung. Nature 373:
393.
Garner, K. J. & Ryder, O. A. (1996) Mitochondrial DNA diversity in
gorillas. Mol. Phyl. Evol. 6: 39-48
Gagneux, P. et al. (2001) Gene flow in wild chimpanzee populations. Phil.
Trans. R. Soc. Lond. Ser. B 356: 889-897
Groves, C. P. (1967) Ecology and Taxonomy of the Gorilla. Nature 213:
890-893
Groves, C. P. (1970) Population systematics for the gorilla. J. Zoology
161: 287-300
Groves, C. P. (2001) Primate Taxonomy. Washington, DC (Smithsonian Inst.
Press)
Groves, J. (2002) Good news for the Nigerian gorillas? Gorilla Journal
24: 12
Grubb, P. (2001) Endemism in African Rain Forest Mammals. In: African
Rain Forest Ecology & Conservation. Weber, W. et al. (eds.). New Haven
(Yale Univ. Press), pp. 88-100
Higuchi, R. et al. (1988). DNA typing from single hairs. Nature 332: 543-546.
Jeffery, K. (2003) Application of Forensic Genetics to the Population
Biology of Western Lowland Gorillas at Lopé, Gabon. Thesis, University
of Cardiff
Jensen-Seaman, M. I. & Kidd, K. K. (2001) Mitochondrial DNA variation
and biogeography of eastern gorillas. Mol. Ecol. 10: 2241-2247
Maley, J. (1996) The African rain-forest - main characteristics of changes
in vegetation and climate change from the Upper Cretaceous to the Quaternary.
Proc. R. Soc. Edin. B 104: 31-73
Morin, P. A. et al. (1994) Kin selection, social structure, gene flow,
and the evolution of chimpanzees. Science 265: 1193-1201
Morin, P. A. et al. (2001) Quantitative polymerase chain reaction analysis
of DNA from noninvasive samples for accurate micro-satellite genotyping
of wild chimpanzees (Pan troglodytes verus). Mol. Ecol. 10: 1835-1844.
Moritz, C. (1994) Defining evolutionary significant units for conservation.
Trends Ecol. Evol. 9: 373-376
Oates J. F. et al. (2002). The cross river gorilla. Gorilla biology. Taylor,
A. & Goldsmith, M. (eds.). Cambridge (Cambridge University Press),
pp. 472-502
Ruvolo, M. et al. (1994) Gene trees and hominoid phylogeny. Proceedings
of the National Academy of Sciences, USA 91: 8900-8904
Taberlet, P. et al. (1996) Reliable genotyping of samples with very low
DNA quantities using PCR. Nucleic Acid Res. 24: 3189-3194
Thalmann, O. et al. (2004) Unreliable mtDNA data due to nuclear insertions:
a cautionary tale from analysis of humans and other great apes. Mol. Ecol.
13: 321-325
Tully, L. A. et al. (2000) A sensitive denaturing gradient gel electrophoresis
assay reveals a high frequency of heteroplasmy in the hyper-variable region
1 of the human mtDNA control region. Am. J. Hum. Genet. 67: 432-443
Vigilant, L. et al. (2004) The problem with gorilla mitochondrial DNA
analysis. Gorilla Journal 28: 15-17