Editor’s Note – This essay was subsequently published as:
Dillon, R.T., Jr. (2019c) Mitochondrial superheterogeneity: What we know. Pp 145 - 153 in
The Freshwater Gastropods of North America Volume 3, Essays on the
Prosobranchs. FWGNA Press, Charleston.
Finally some decent sample sizes, thank heaven! I remember thinking this to myself when the
abstract of the recent paper in Zoologica Scripta by Nathan Whelan and Ellen
Strong fell upon my eyes in January [1]. Perhaps
now, after twenty years of bouncing around on the fringes of evolutionary
science, the phenomenon we here define as “mitochondrial superheterogeneity”
will finally attract the attention it so richly deserves.
A population can be said to demonstrate “mitochondrial
superheterogeneity” when two or more of its members demonstrate 10% sequence
divergence or greater for any single-copy mtDNA gene, not sex-linked. The phenomenon was first documented (as far
as I am aware) in 1996 by Thomaz, Guiller and Clarke in European populations of
the land snail, Cepaea nemoralis [2]. It
has now been reported at least three additional times in land snails [2], twice
in populations of freshwater pulmonates [3], once in oriental pomatiopsids [4],
three times in Old World cerithiaceans [5] and six times in populations of
North American pleurocerids [6, 7]. But
the new study by Whelan & Strong is by far the most thorough.
W&S sampled 20 individuals from each of 12 populations of
pleurocerid snails collected from North Alabama, sequencing three genes from
all 240 individuals [8]: (mitochondrial) COI, (mitochondrial) 16S rRNA, and
(nuclear) histone H3. Six of the
populations sampled were of Leptoxis (identified as praerosa in Tennessee
drainages or ampla in Alabama drainages) and six of the populations were of Pleurocera
(identified as pyrenella in Tennessee drainages or prasinata in Alabama
drainages). The specific identities of
these populations are a complicated question, which we will defer to a later
essay. But here are the important
findings:
For example, the COI results obtained by W&S from five
of their Leptoxis populations are depicted above. The 19 snails [8] sequenced from the Shades
Creek population (shown in dark blue) had three strikingly different
haplotypes: N=16 in the modal category, N=2 about 68 mutational steps (9%)
different from the modal, and N=1 about 88 steps (12%) different. Notice that the red intermediate points in
the figure below are mostly quite distant from the branch tips and entirely
hypothetical.
(2) Peculiar patterns of interpopulation
haplotype sharing. Let us focus on
that single Shades Creek snail demonstrating a 12% COI sequence divergence from
its N = 16 associates, marked with a red star in the figure above. That CO1 sequence (KT163940) is only 1%
different from the sequence demonstrated by N=17 of the 20 snails sampled from
the Cahaba River at the US52 bridge, 5 km downstream and then back up the main
river about 20 km (e.g., KT164013).
(3) Tight coupling in the evolution and
distribution of separate mitochondrial genes. W&S confirmed that individuals bearing
highly divergent COI sequences also tend to bear highly divergent 16S sequences. So the starred individual in the COI network above also demonstrated a 16S sequence rather strikingly divergent from the modal
Shade Creek 16S sequence. And that
divergent 16S sequence also very nearly matched the modal sequence recovered
from the Cahaba River at the US52 bridge [9].
Such results were first reported in 2004 by Dillon &
Frankis [5], working with Pleurocera proxima in southern Virginia [10]. Our 2004 mtDNA sample sizes were very small –
just three individuals from three P. proxima populations, plus single
individuals of P. catenaria dislocata and P. semicarinata. But we did sequence both the COI and 16S
genes, giving us a total of 11 sequences for each gene. All three of our P. proxima from Naked Creek matched
in both sequences, and all three of our P. proxima from Cripple Creek matched
in both sequences. But the individuals
we sampled from Nicholas Creek demonstrated superheterogeneity: the CO1
sequence of the (N=2) mode and the (N=1) minority differing by 16.9%. And that same N=1 minority snail also
differed from the (N=2) modal 16S sequence in Nicholas Creek by 14.1%.
(4) No evidence of cryptic speciation. Although the sample sizes of mtDNA sequence
data in all three of my previously-published works on this subject have been
quite small, all three have focused on pleurocerid populations for which I have
tested Hardy-Weinberg equilibrium at multiple polymorphic allozyme loci with
sample sizes of at least N = 30, and often much larger. These are not weak tests, and if cryptic
species had been present, I have reason to think that I would have found them
[11]. But I have found no evidence of
non-random mating in any population from which I have reported mitochondrial
superheterogeneity [12].
Whelan & Strong have reinforced this result with some classical
(i.e., entirely qualitative) studies of shell and radular morphology, pallial
oviduct anatomy, and head-foot coloration in their 12 pleurocerid populations
from North Alabama. Their morphological
observations are nicely illustrated and described in quaint, Victorian style
(e.g., “Outermost denticle on each side delicate, flimsy, variable in position”). Although W&S do note some intrapopulation
variation in some morphological traits, none seemed to correspond with mtDNA
clade.
(5) No evidence of NUMTs. In some circumstances mitochondrial genomes
or portions of genomes have been discovered transposed into the nucleus, yielding
“NUclear MiTochodrial DNA” which, being “free to evolve,” can diverge greatly
from the bona fide mtDNA genome. W&S
used a next-generation technique to sequence the entire mitochondrial genome of
a Leptoxis individual from which they had obtained highly-divergent COI and 16S
sequences with their standard (Sanger) methods.
The next-generation sequences matched the Sanger sequences, counter to
the NUMT hypothesis.
(6) No evidence of DUI. Doubly-uniparental inheritance is the surprising
situation where two highly-divergent mtDNA genomes have evolved within a single
species, one transmitted through eggs and the other transmitted through sperm. The phenomenon has been documented in a great variety of bivalve groups, including the freshwater unionids, but is
unknown in gastropods. And indeed,
W&S could find no systematic differences in the mtDNA genomes isolated from
the male and female samples of pleurocerids they analyzed.
So that’s what we know about mitochondrial
superheterogeneity as of March 2016, and what doesn’t cause it. But what, then, might a reasonable
explanation be? Are data such as these rudely
yelling something important to us, in some language we do not understand? Tune in next time!
Notes
[1] Whelan, N.V. & E. E. Strong (2016) Morphology, molecules and taxonomy: extreme
incongruence in pleurocerids (Gastropoda, Cerithiodea, Pleuroceridae).
Zoologica Scripta 45: 62 – 87. Open
Access: [html]
[2] Thomaz, D., Guiller, A. & Clarke, B. (1996) Extreme
divergence of mitochondrial DNA within species of pulmonate land snails.
Proceedings of the Royal Society of London B: Biological Sciences, 263,
363–368. Goodacre, S. L. & C. M.
Wade (2001) Patterns of genetic
variation in Pacific island land snails: the distribution of cytochrome b
lineages among Society Island Partula.
Biol. J. Linn. Soc. 73: 131-138. Guillier,
A. et al. (2001) Evolutionary history of
the land snail, Helix aspersa in the western Mediterranean: preliminary results
inferred from mitochondrial DNA sequences.
Molecular Ecology 10: 81-87. Haase, M. et al. (2003) Mitochondrial differentiation in a polymorphic
land snail: evidence for Pleistocene survival within the boundaries of permafrost. J. Evol. Biol. 16: 415-428.
[3] My good friend and colleague Amy Wethington documented a
striking case of mitochondrial superheterogeneity (both COI and 16S) in a
Charleston population of Physa acuta in 2003.
We never could get a separate paper through the reviewers, but the data
were published in her (2004) dissertation.
Andrea Walther and colleagues had better luck with the reviewers in her
(2006) paper on Laevapex fuscus. For
more, see:
- Phylogenetic Sporting and the Genus Laevapex [20July07]
[5] Facon, B. et al.
(2003) A molecular phylogeography
approach to biological invasions of the New World by parthenogenetic thiarid
snails. Molecular Ecology 12:
3027-3039. Lee, T., H. C. Hong, J. J.
Kim & D. O’Foighil (2007)
Phylogenetic and taxonomic incongruence involving nuclear and mitochondrial
markers in Korean populations of the freshwater snail genus
Semisulcospira. Molecular Phylogenetics
and Evolution 43: 386-397. Miura, O., F.
Kohler, T. Lee, J. Li, & D. O’Foighil (2013) Rare, divergent Korean Semisulcospira spp.
mitochondrial haplotypes have Japanese sister lineages. J. Moll. Stud. 79: 86 – 89.
I reviewed the (2007) paper by Taehwan Lee and his
colleagues on this blog in February, 2008:
- Semisulcospira II: A second message from The East [1Feb08]
I reviewed the Dillon & Robinson (2009) results in my
blog post of March, 2009. See:
- The Snails The Dinosaurs Saw [16Mar09]
Dillon, R. T. Jr, and J. D. Robinson (2016) The hazards of DNA barcoding, as illustrated
by the pleurocerid gastropods of East Tennessee. Ellipsaria 18: 22-24. [PDF]
More about this little work in a future post.
[8] Well, to be
precise, one of the 240 snails was excluded by misidentification. Then no COI sequence was ultimately obtained
for 13 individuals, no 16S sequence obtained for 16 individuals, and no H3
sequence obtained for 15 individuals.
But we can certainly call it 240 for round numbers.
[9] The Genbank
accession number of the “clade 14” Shades Creek 16S sequence is KT164167. Blasting that sequence against the 16S
sequence from a random clade 15 Shades Creek snail (KT164165) yielded an
identity of 94%. Blasting KT164167
against a random Clade 14 16S sequence from the Cahaba River at US52 (KT164235)
yielded an identity of 99%.
[10] In the first paragraph of their discussion, our good
friends Nathan Whelan and Ellen Strong wrote, “We show for the first time in
pleurocerids that there is complete congruence in phylogenetic signal between
the COI and 16S mitochondrial genes.” They have completely neglected to cite
the Dillon & Frankis (2004) paper, either here or anywhere else in their
work, even though Dillon & Frankis was heavily-cited by Dillon &
Robinson (2009), of which they obviously were aware. I am sure this was just a simple oversight.
[11] In fact, we did find a cryptic species in the same
(2011) allozyme survey that forms the basis of the recent note by Dillon &
Robinson [7]. Put another bookmark here
– we’ll come back to this later, as well.