ABSTRACT:

1
Discovery of a free-living picocyanobacterium
with a hybrid ribosome and chlorophyll d as
the major light-harvesting pigment. S.
Miller1,2, and A. M. Wood1, 1Center for Ecology
and Evolution University of Oregon, Eugene OR,
USA 2Division of Biological Sciences, The
University of Montana, Missoula MT 59812
ABSTRACT In the course of a biotic inventory of
the Salton Sea (California), we isolated a
unicellular picocyanobacterium with unusual
lime-green pigmentation (Fig. 1A) due to the
presence of chlorophyll d as the predominant
light-harvesting pigment (Fig. 2). Fine
structure reveals many peripheral thylakoids and
no phycobilisomes (Fig. 3) although we find
evidence of PE genes (Everroad, not shown). The
strain is closely related to Acaryochloris
marina, which was isolated as a symbiont from a
colonial ascidian, and which is found in pristine
oceanic waters. In contrast, the free-living form
we discovered lives in a hypereutrophic high
salinity environment. Together, the strains
represent a new lineage within the cyanobacterial
radiation (Fig. 4). The V1 region of the
ribosome contains an insert not found in any
other cyanobacteria (Fig. 5) and which appears to
have been acquired from a proteobacterial donor
between 10 and 200 myaBP (Fig. 5). These results
extend the concept of a mosaic prokaryotic genome
that can be fluid across great evolutionary
distances to our most widely used phylogenetic
marker.
0.5 ?m
DATING THE GENE TRANSFER -- Probably occurred
10-200 Ma ago, in contrast to 2200 MA ago, when
the ?-proteobacteria and cyanobacteria last
shared a common ancestor. The origin of the
insert in the V1 spur is constrained to the time
period after these strains last had a common
ancestor with Synechococcus IR11, but prior to
their divergence from each other. We used ML to
implement a general time reversible nucleotide
substitution model with among site
heterogeneity, with and without an imposed clock.
For the entire tree in Fig. 4, as well as every
possible sub-tree defined by the internal nodes
in the phylogeny., pairs of clock and no-clock,
models were compared to the likelihood ratio
tests. While the tree as a whole does not obey a
global molecular clock, sequence evolution of the
Node B Clade does conform to a local clock (-2XL
0.13, P0.72). We next estimated the timing
of the LGT event using the fossil record
calibrated 16S SSU tRNA clock for a group of
bacterial endosymbionts of aphids (Buchnera
spp.), which gives a rate of 0.01-0.02 nucleotide
substitutions per site per 50 million years.
This provides an estimate of the timing of the
divergence between Node A and B at between 10 and
200 Ma ago. This stands in sharp contrast to the
time, greater than 2200 Ma ago, that the
cyanobacteria and b-proteobacteria last shared a
common ancestor.