PCP Gene Family in Symbiodinium from Hippopus hippopus: Low Levels
of Concerted Evolution, Isoform Diversity, and Spectral Tuning
of Chromophores
Jay R. Reichman,* Thomas P. Wilcox,* and Peter D. Vize *School of Biological Sciences, University of Texas at Austin; and Department of Biological Science, University of Calgary,
Calgary, Canada
Photosynthetic dinoflagellates have evolved unique water-soluble light harvesting complexes known as peridininchlorophyll a–binding proteins (PCPs). Most species of dinoflagellates express either 14 to 17 kDa or 32 to 35 kDa
mature PCP apoproteins and do so in stable combinations of isoforms that differ in isoelectric point (pI). The source
(posttranslational modification, protein degradation, or genetic) and functional significance of PCP isoform variation
have remained unclear. PCPs are encoded by multigene families. However, previous reports conflict over the diversity of
PCP genes within gene arrays. We present the first genomic characterization of the PCP gene family from a symbiotic
dinoflagellate. Symbiodinium from the Pacific bivalve Hippopus hippopus (203) contains genes for 33 kDa PCP
apoproteins that are organized in tandem arrays like those of free-living dinoflagellates Amphidinium carterae,
Lingulodinium (Gonyaulax) polyedra, and Heterocapsa pygmaea. The Symbiodinium 203 PCP cassette consists of
1,098-bp coding regions separated by approximately 900-bp spacers. The spacers contain a conserved upstream sequence
similar to the promoter in L. polyedra. Surprisingly, sequences of cloned coding regions are not identical, and can differ
at up to 2.2% of the nucleotide sites. Sequence variation is found at both silent and nonsilent sites, and analysis of cDNA
clones indicate that the variation is present in the mRNA pool. We propose that this variation represents nucleotide
diversity among PCP gene copies that are evolving under low-level concerted evolution. Interestingly, the predicted
proteins have pIs that are within the range of those published for other species of Symbiodinium. Thus, posttranslational
modifications are not necessary to explain the multiple PCP isoforms. We have also identified several polymorphic sites
that may influence spectral absorption tuning of chromophores.
Introduction
Photosynthetic dinoflagellates have evolved unique
extrinsic light-harvesting complexes known as peridininchlorophyll a–binding proteins (PCPs). PCPs apparently
function within the aqueous lumen of dinoflagellate
chloroplasts and are distinct in terms of their combination
of pigments and water solubility (Haxo et al. 1976;
Prezelin and Haxo 1976). Additionally, they occur solely
in dinoflagellates and do not share significant amino acid
sequence homology with other chromophore-binding
proteins (Triplett et al. 1993; Norris and Miller 1994),
including phycobiliproteins or membrane-bound, lightharvesting complexes.
Within chloroplasts, PCPs allow harvesting of bluegreen 435 to 550 nm light by peridinins and efficient
transfer of energy to chlorophyll a (Song et al. 1976;
Siegelman, Kycia, and Haxo 1977; Larkum 1996; Moffat
1996; Damjanovic, Ritz, and Schulten 2000). The first
high-resolution crystal structure of PCP from the freeliving dinoflagellate Amphidinium carterae (Hofmann
et al.1996) indicates that tuning of this transfer is achieved
by the specific physical arrangement of the chromophore
complex and the surrounding amino acids. In addition,
changes in the polarity of the PCP protein environment
neighboring the furanic rings and polyene chains of
peridinins could modify the spectroscopic properties of
these accessory pigments (fig. 1).
Key words: peridinin, PCP, gene family, dinoflagellate, Symbiodinium, concerted evolution.
E-mail: [email protected].
Mol. Biol. Evol. 20(12):2143–2154. 2003
DOI: 10.1093/molbev/msg233
Molecular Biology and Evolution, Vol. 20, No. 12,
Ó Society for Molecular Biology and Evolution 2003; all rights reserved.
There are two general size classes of PCP apoproteins, commonly called short and long PCPs:14 to 17
kDa monomers/homodimers and 31 to 35 kDa monomers.
Most dinoflagellate species express only one size class of
apoproteins, although a few express both (Prezelin and
Haxo 1976; Govind et al. 1990). The axis of symmetry in
the amino acid sequence of long PCPs strongly suggests
that the original gene for long PCPs arose from a duplication and fusion event between genes for short PCPs (Le
et al. 1997), and analysis of nucleotide sequences supports
this hypothesis (Hiller et al. 2001).
Regardless of which size PCP an individual dinoflagellate species expresses, it does so in multiple
isoforms with distinct isoelectric points (pIs). Stable
combinations of isoforms are species specific (Haxo et al.
1976; Prezelin and Haxo 1976; Siegelman, Kycia, and
Haxo 1977; Chang and Trench 1982, 1984; Trench and
Blank 1987). There are differences in spectroscopic
properties of isoforms between species and among
isoforms within species (Haxo et al. 1976; Prezelin and
Haxo 1976; Song et al. 1976; Iglesias-Prieto, Govind, and
Trench 1991). The source of PCP isoform variation, be it
posttranslational modification, protein degradation, or
genetic, has remained unclear (Haxo et al. 1976; Siegelman, Kycia, and Haxo 1977; Chang and Trench 1984;
Triplett et al. 1993; Ogata et al. 1994; Sharples et al. 1996;
Hiller et al. 2001). Likewise, the functional significance of
isoform diversity has yet to be resolved.
With the exception of descriptions of cDNAs from
two Symbiodinium species, all other details on PCP gene
structure and organization are based on three free-living
dinoflagellates: A. carterae, Lingulodinium (Gonyaulax)
polyedra, and Heterocapsa pygmaea (Triplett et al. 1993;
Norris and Miller 1994; Sharples et al. 1996; Le et al.
2143
2144 Reichman et al.
FIG. 1.—(a) A. carterae trimer of 35-kDa PCPs (NCBI PDB ID 1PPR [Hofmann et al. 1996]). These monomers typically contain (b) eight
peridinin, (c) two chlorophyll a, (d ) two digactosyl diacyl glycerol, and (e) combined chromophores complex. Regions of interaction between peridinin
and amino acids with polar side chains within PCPs include ( f ) furanic ring and (g) polyene chain of peridinin.
1997; Hiller et al. 2001; Weis, Verde, and Reynolds 2002).
The nuclear genes that encode PCPs are intronless and
exist as multigene families set in tandem arrays. However,
previous reports conflict over the diversity of genes within
these arrays. Evidence has mounted to suggest that PCP
gene families may not be highly conserved in general and
that the expression of distinct PCP isoforms is primarily
due to genetic diversity among gene copies in PCP arrays
(Triplett et al. 1993; Sharples et al. 1996; Le et al. 1997;
Hiller et al. 2001). If so, the expectation that genes in
tandem arrays evolve in concert may not strictly be true for
PCP genes. Interestingly, nucleotide diversity has also
been reported in other dinoflagellate multigene families,
including those for luciferin-binding protein and Rubisco
(Lee et al. 1993; Machabee’, Wall, and Morse 1994;
Rowan et al. 1996).
Detailed information about PCP gene families from
Symbiodinium is needed to lay the groundwork for
evolutionary analyses of PCP genes across dinoflagellate
genera and species. Furthermore, the issue of genetic
diversity contributing to expression of multiple isoforms in
a functional manner also needs to be addressed. We
present the first genomic characterization of the PCP gene
family from a symbiotic dinoflagellate, Symbiodinium
from the Pacific bivalve Hippopus hippopus (RK Trench
culture collection number 203; clade C, sensu Rowan and
Powers 1991a, 1991b, 1992; Wilcox 1998; LaJeunesse
and Trench 2000; LaJeunesse 2001). Symbiodinium 203
PCP gene structure, organization, and copy number are
compared with genes from L. polyedra. The presented data
uncover considerable PCP gene diversity in Symbiodinium
203 and demonstrate how this diversity acts as a primary
source of variability in PCP isoforms. Amino acid
substitutions are mapped onto the A. carterae PCP crystal
structure to consider potential functional significance of
variable residues, especially with regard to how polymorphic sites may influence the spectral tuning of
peridinins.
Materials and Methods
Algal Cultures
Dr. Robert K. Trench kindly donated Symbiodinium
203 from his culture collection at University of California
Santa Barbara. Uni-algal subcultures were maintained in
0.45 mm filtered Provasoli’s Enriched Seawater (PES) or
Guillard’s F/2 (Sigma, Inc.). The cultures were grown
under full-spectrum fluorescent light banks (Sylvania
40W 4100K Designer) at 80 l mole quanta/m2/s on
a 12:12 light-to-dark cycle at 278C. Cultures were serially
transferred every 3 to 4 weeks. L. polyedra (CCMP
number 1738) was obtained from the Provasoli – Guillard
National Center for Culture of Marine Phytoplankton.
Subcultures of L. polyedra were grown under conditions
similar to Symbiodinium 203 except that L. polyedra was
only grown in F/2 and the temperature was maintained at
218C.
PCP Gene Family Evolution in Symbiodinium 2145
Table 1
Primers Used for Amplification and Sequencing
Table 2
Primer Sets, Annealing, and Extension Parameters
Upper/Forward Lower/Reverse
Primers
Primers
Reaction
U325
L537
U448
L423
U935
L(67)
U(1540)
L24
U3
L913
U(28)
L(1180)
SYBRf938
SYBRr1006
Primer Sequence
(59–39)
AAGAATTCGAAGGACGCAGCA
GAAGC
CAGAATTCCTTCATGTACGCTGG
CAC
TCGGTCCCCAAAGCAAAGGTCA
CATTCACGGCATCCCAGTCAGC
CCGTGCCCAAGTCCACTGTCA
GCAGGATGATTGGGATGAGT
GAGCCGAACACATCCAGCAG
AGCTTTCCTTGCTCCACGCAC
GGTGCGTGGAGCAAGGAAAG
GCATTCACAGCTTCGTAGT
TCCGGCCCACTTTTAGTTTT
TTTTCCCATTTGTTTCAGAG
TGCCCAAGTCCACTGTCATG
TGTTGGTCACGGTGGAATCA
NOTE.—Numbers indicate location of 59 end of primer relative to the
Symbiodinium 203 coding region. Those in parentheses are outside of the coding
and variable due to indels in the spacers.
Nucleic Acid Extractions
Total genomic DNA was extracted either by a method
previously developed for symbiotic dinoflagellates (Rowan
and Powers 1991b) or by a modification of the DNAeasy
tissue extraction kit (Qiagen, Inc.) 1999 protocol. Steps
one and two of the Qiagen protocol were replaced by the
following: Algal pellets were resuspended in 500 ml of
standard 2X CTAB buffer (Coffroth et al. 1992).
Resuspended cells were ground within microcentrifuge
tubes and then 4 ml of 10 mg/ml Proteinase K was added
to each tube. The tubes were incubated at 658C for 2 h and
were inverted every 30 min. The extractions then carried
forward from step three of the Qiagen protocol.
Nucleic acid extractions done by the Rowan and
Powers (1991b) method also contained substantial
amounts of usable RNA. To purify the RNA for reverse
transcription PCR (described below), total nucleic acid
preps were diluted 1:20 in 1X DNase buffer and incubated
with DNase I (Ambion, Inc.) at 378C for 1 h. The RNA
was phenol/chloroform extracted, ethanol precipitated,
desalted, air-dried, and resuspended in RNase-free ddH2O.
PCR
PCR
PCR
Reverse
transcription
cDNA SYNTH
Cycle SEQ
Auto SEQ
Quantitative
real-time PCR
Primer(s)
Annealing
Extension
U325/L537
U448/L423
U(-28)/L(1180)
OLIGOdT
1 min at 508C
1 min at 558C
45 s at 508C
1 min at 308C
U3/L913
Various
Various
SYBRf938/
SYBRr1006
30 s at 558C
30 s at 428C
5 s at 508C
20 s at 608C
45 s at 728C
2 min 15 s at 728C
1 min 15 s at 728C
30 min rise 308C
to 658C
1 min at 728C
1 min at 708C
4 min at 608C
20 s at 728C
PCR Conditions, Identification, and Purification of
Products
Standard PCR conditions (Palumbi 1996) were used
with modification to the 10X PCR buffer (200 mM TrisHCl pH 8.8, 100 mM KCl, 100 mM (NH4)2SO4, 20 mM
MgSO4, 1% Triton X-100, and 1 mg/ml BSA). Thermal
cycle profiles were adjusted to accommodate the annealing
temperature of the primer sets and the length of the
expected PCR product (table 2). Each round of PCR
included a negative control to check for contamination of
reagents. When multiple bands were present in a given
reaction, Southern hybridization (Sambrook 1989) identified PCP gene amplification products. PCR products were
transferred to Hybond-Nþ nylon membranes (Amersham,
Inc.). Probes were created with DECAprime II random
priming DNA labeling kits (Ambion, Inc.). Gel excised
PCR products of interest were purified with QIAEX II gel
extraction kits (Qiagen, Inc.)
Reverse Transcription PCR
A BcaBEST RNA PCR kit version 1.1 (TaKaRa,
Inc.) was used for reverse transcription and amplification
of cDNA. The oligo dT primer included in the kit was used
for synthesis of single-stranded cDNA following a thermoprofile suggested by the manufacturer (table 2). Subsequent amplification of double-stranded cDNA was done
with the U3/L913 primer set using ‘‘A Method’’ from the
kit manual. Each round of reverse transcription PCR
included controls to check for contamination.
Primer Design
Primer used for standard PCR, reverse transcription
PCR, and quantitative real-time PCR and for sequencing
are listed in table 1. Published PCP gene sequences
Symbiodinium from Acropora formosa, A. carterae and L.
polyedra were aligned with ClustalX to identify conserved
regions and initial design of primer set U325/L537 (Norris
and Miller 1994; Sharples et al. 1996; Le et al. 1997;
Thompson et al. 1997). Thereafter, sequences from
derived clones were used to design primers. Primer
selection for standard and reverse transcription PCR was
optimized with Oligo 4.0 (National Biosciences, Inc.).
Primers for quantitative real-time PCR were designed with
Primer Express (Applied Biosystems, Inc.).
Cloning and Screening Plasmid Libraries
The PCR product from U325/L537 was blunt cloned
into the pBluescript II SK vector (Stratagene, Inc.). All
other PCR products were cloned into pCR2.1 vectors with
TA cloning kits (Invitrogen, Inc.) according to the
manufacturer’s protocol. Cells from transformed colonies
were lysed by boiling in ddH2O and were screened by
PCR to detect inserts. Standard minipreps were created
for each positive clone. Plasmid DNA was extracted and
purified with QIAamp DNA Mini kits (Qiagen, Inc.).
PCR products from these clones were gel excised,
purified as described above, and eluted in ddH2O for
sequencing.
2146 Reichman et al.
Sequencing and Sequence Analysis
All published sequences for this project were
sequenced at least twice and were done in both the
forward and the reverse directions. The sequencing
annealing and extension parameters are described in table
2. The clone from the U325/L537 PCR product was
sequenced with an fmol DNA Cycle Sequencing System
kit (Promega, Inc.). Automated sequencing was done for
all other clones using BigDye (version 2) Terminator kits
(ABI, Inc.), and data was collected on a PerkinElmer ABI
PRISM 377 DNA Sequencer. Sequence contigs were
assembled with Seqman (DNASTAR, Inc). Translation of
predicted proteins and pI estimation with done with Gene
Runner version 3.04 (Hastings Software Inc.) and with
Edit Seq (DNASTAR, Inc.). GenBank database searches
for similar nucleotide and amino acid sequences were done
using the Blast algorithm (Altschul et al. 1990). Nucleotide
sequences of the clones and an alignment (described
below) were submitted to GeneBank under accession
numbers AY149122 to AY149139.
PCR Recombination and Fidelity Controls
Three unique clones, amplified from tandem coding
regions, were combined in equal quantities and used as
templates for PCR reactions that were run under conditions
described above. Amplification products from each reaction were TA cloned. Three sets of 16 subclones were
isolated and then sequenced with U448 and L423. Each
was then identified as being either one of the original
templates or a recombinant. The location of the recombination was determined to be in either the coding
region or the spacer, and then the observable recombination frequency was calculated as a percentage.
In addition, a complete coding sequence (cds) clone
was amplified in five separate PCR reactions under the
same conditions as above. A subclone was isolated from
each reaction and sequenced with U(28), U448, L913,
and L(1180) to check for nucleotide substitutions introduced by the PCR process. The maximum Taq error rate
that would produce the observed number of mutations was
estimated. A standard likelihood analysis assuming a binomial distribution was conducted to determine the 95%
confidence intervals for the Taq error rate. A comparison
was then made to the substitution rate observed within the
region common to all genomic complete cds and cDNA
clones.
Nucleotide Sequence Divergence
K-Estimator version 5.5 (Comeron 1995, 1999) was
used to calculate the number of nonsynonymous (Ka) and
synonymous (Ks) nucleotide substitutions per site for all
paired comparisons of complete coding regions. No
comparison restrictions were introduced, and all base pair
sites within 365 codons were analyzed for each comparison. The Kimura two-parameter method was used to
correct the number of substitution hits per site. Ka and Ks
values were generated from separate analyses, and then
Ka/Ks was calculated for each pair of compared sequences.
DNA coding regions were aligned with ClustalX (Thomp-
son et al. 1997) and MegaAlign (DNA Star, Inc.). To
visually depict the nucleotide sequence divergence, we
used PAUP* version 4.0b10 (Swofford 1998) to generate
a Neighbor-Joining tree from Kimura two-parameter DNA
distances.
PCP Gene Copy Number and Genome Size Estimation
The number of PCP genes per genome for Symbiodinium 203 was calculated by the overall equation:
PCP genes 6 rgenes
pg genomic DNA 6 rpg
3
pg genomic DNA
genome
PCP genes 6 rgenes
¼
genome
ð1Þ
where r equals the standard deviation.
The (PCP genes 6 rgenes)/(pg genomic DNA) term
in equation 1 was determined by quantitative real-time
PCR with a PerkinElmer ABI 7700. Data were analyzed
with ABI Sequence Detection Systems software version
1.7. Amplifications of a 69-bp PCP gene segment were
compared between known amounts of Symbiodinium 203
genomic DNA (10,000 pg) and a dilution series of
linearized PCP clone (1,492,000 fg to 14.92 fg in 10-fold
dilutions). Concentrations of genomic and plasmid DNA
were quantified in triplicate on a Beckman Coulter DU640
spectrophotometer. A 1.492 lg sample of the clone was
digested with two units NotI (New England Biolabs, Inc.)
at 378C for 1 h to cut the pUC19 vector at a single site
outside the insert. Duplicate reactions were set up for each
of the standards and for negative controls. Triplicate
reactions were set up for the Symbiodinium 203 genomic
samples. Reagents were assembled in master mix and were
distributed so that each reaction contained 0.4 mM dNTPs,
10.0 mM Tris-HCl pH 8.3, 50.0 mM KCl, 6.0 mM MgCl2,
1.0% glycerol, 0.01% Tween 20, and 1:4000 dilution of
SYBR Green I (Molecular Probes, Inc.), two units of
SuperTaq (Ambion, Inc.), 0.1 lM SYBRf938 primer,
0.1 lM SYBRr1006 primer, and the DNA template.To
compare PCP copy number to ng of genomic DNA, PCP
gene copies/fg of PCP clone were converted as follows:
ðplasmid þ insertÞ 1 mole bp 6:0231023
1g
3
3
3 15
5108 bp
660 g
10 fg
mole
179ðplasmids þ insertsÞ
ð2Þ
¼
fg
The (pg genomic DNA 6 rpg)/(genome) term in
equation 1 was estimated by comparing the average
genome sizes of Symbiodinium 203 and L. polyedra by
flow cytometry using a Beckton Dickenson FACS Calibur.
The instrument was equipped with a 15 mW Argon laser
producing excitation at 488 nm. A modified version of the
Veldhuis, Cucci, and Sieracki (1997) protocol for fixing
and staining the cells was used. Five hundred microliters
of algal cells suspended in F/2 was combined with an
equal volume of 4% paraformaldehyde and incubated at
228C for 15 min, after which Triton X-100 and RNase-A
were added to concentrations of 0.0033 % (v/v) and 1 ng/
ml, respectively. The solutions were mixed by inversion
PCP Gene Family Evolution in Symbiodinium 2147
and kept at room temperature for 20 min. Cells were
pelleted by centrifugation at 3000 rpm at 48C. After
decanting the liquid, the fixed cells were resuspended in
500 ll of TE. Triton X-100 concentration was adjusted to
0.002 % (v/v) and 20X PicoGreen (Molecular Probes, Inc.)
was added to achieve a final 2X staining concentration.
Cells and reagents were mixed by inversion. Cells were
stained at 228C for 1 to 2 h before flow cytometry.
PicoGreen Fluorescence of DNA was measured through
the FACS FL1 filter at 530 6 30 nm. The DNA content of
Symbiodinium 203 nuclei was then estimated as the ratio
of fluorescence from Symbiodinium 203 and L. polyedra
times the estimated DNA content of 200 pg/cell for L.
polyedra (Holm-Hansen 1969; Spector 1984). The methods from Melissinos (1966) were used to accurately
account for the propagation of indeterminate errors.
Amino Acid Substitution Modeling
Predicted amino acid substitutions were mapped onto
the crystal structure for the A. carterae PCP trimer 1PPR
(Hofmann et al. 1996) using Swiss-PDB Viewer version
3.7(b2) (Glaxo Wellcome, Inc.) as follows: Pairwise amino
acid alignments between Symbiodinium 203 and A.
carterae PCPs were created to identify conserved and
variable sites. Mutations were individually introduced in
the 1PPR structure at each of the fixed and polymorphic
sites. As substitutions were made, rotomer conformations
were optimized by the software. Distances were then
calculated between polar side-chain residues in polymorphic sites and furanic rings and/or polyene chains of
peridinins within the same monomer. A single-layer pdb
was rendered, reflecting all changes. Three-dinemsional(??) structure images of selected molecules from
the holoprotein were rendered using POV-Ray version 3.1
and version 3.5 (POV-Ray Team). Residue numbers
within renderings were based on the 1PPR structure.
Results
Organization and Diversity of Genomic Coding and
Spacer Regions
The cloning strategy for this project is summarized in
figure 2. Long PCP coding regions of Symbiodinium from
Acropora formosa (L13613), A. carterae (Z50792, and
Z50793), and L. polyedra (U93077) shared 70.4% to 90.8%
identity and consequently when aligned did not produce
large, conserved blocks that could be easily targeted for
amplification. However, primers U325 and L537 based on
sequence from the 59 half of L13613 successfully amplified
a 212-bp fragment from Symbiodinium 203 (fig. 2a). The
sequence of the small cloned fragment was identified by
a Blast search as partial PCP gene.
PCP genes from the free-living dinoflagellates A.
carterae, L. polyedra, and H. pygmaea were previously
shown to exist in tandem arrays (Sharples et al. 1996; Le
et al. 1997; Hiller et al. 2001). Although it was not known
whether Symbiodinium 203 had long or short PCP genes,
we assumed that either variety would also be arranged in
tandem. Sequence from the 212-bp fragment was used to
design outward facing primers U448 and L423 to amplify
between adjacent gene copies. Amplification products of
multiple sizes were present in the reaction. A genomic
Southern blot probed with the small PCP fragment
produced hybridization to a 1.9-kb band (not shown) that
was subsequently excised and purified. Direct sequencing
of the 1.9-kb PCR product contained many unresolved
bases, suggesting that multiple templates were present in
the reaction. The 1.9-kb fragments were TA cloned, and
a single clone was completely sequenced by primer
walking. Open reading frames at the 59 and 39 ends of
this clone both translated to PCP and flanked a 903-bp
spacer region. This sequence confirmed that the PCP genes
from Symbiodinium 203 are arranged in tandem arrays like
those from free-living dinoflagellates.
Sequencing additional 1.9-kb clones revealed substitutions among gene copies distributed throughout the
coding and spacer regions (fig. 2b). As expected, the
spacers contained the majority of the substitutions,
including insertions and deletions. Despite the highly
variable nature of the spacers, each contained a conserved
13-bp sequence CTTGAATGCAGAA, approximately
201 to –188 bp upstream of the start codon. Nine of
the bases in this sequence are identical to and in the same
relative location as the promoter previously identified from
the L. polyedra luciferase and PCP spacers (Li and
Hastings 1998). This conserved sequence is very likely to
be the Symbiodinium 203 PCP gene promoter.
Spacer sequence was also used to design primers that
flank coding regions. Amplification of Symbiodinium 203
genomic DNA with inward-facing primers U(28) and
L(1180) produced a 1.1-kb band. Because direct sequencing of the PCR product indicated that multiple templates
were again present, these products were cloned and
individually sequenced. Of the 11 clones sequenced, 10
contained a 1,098-bp complete, long PCP coding sequence
(fig. 2c). One clone contained a truncated coding region
with a stop codon occurring 27 bases earlier than others.
All but two of the complete coding regions were distinct
from each other in terms of their nucleotide substitutions
(97.8% to 99.5% identical). The majority of nucleotide
substitutions between Symbiodinium 203 complete coding
regions were in synonymous codon positions resulting in
Ka/Ks , 1 (mean 0.29 6 0.18). Only a single pairwise
comparison had a Ka/Ks . 1 (¼ 1.11).
Low PCR Recombination Frequency and
High PCR Fidelity
To control for possible generation of recombinant
DNA sequences in amplifying genes arrayed in tandem
(Bradley and Hillis 1997), a reamplification and subcloning experiment was conducted using pairs of distinctive
1.9-kb clones as the templates. All other PCR conditions
were the same as used on Symbiodinium 203 genomic
DNA. Sixteen subclones from each of the amplifications
were partially sequenced to identify them as either one of
the original templates within the PCR reaction or as
recombinant. The first 500 bp of sequence from both the
59 and the 39 end were used for the identification. Four of
the 48 subclones had different 59 and 39 identities. For
each of these recombinants, the 59 and 39 coding regions
2148 Reichman et al.
FIG. 2.—Cloning strategies. Horizontal bars represent coding regions. Horizontal lines represent spacer. Smaller black vertical bars represent
hypothetical locations of substitutions; open vertical bars represent putative promoter. (a) Original clone, (b) clones between adjacent coding regions,
(c) clones of complete coding regions, and (d ) cDNA clones.
sequenced completely matched one or the other template,
suggesting that recombination had likely occurred in the
spacer. One of the subclones was excluded as a contaminant. The observable recombination frequency was 8.5%.
The maximum rate at which Taq error introduced
substitutions in the PCR reactions was also estimated.
There were no changes introduced into any of the five
1,198-bp subclones amplified from the same template in
separate reactions. For the 5,990 bp sequenced, the
maximum error rate that would produce zero substitutions
is approximately 5 3 104 errors/bp with a 95% likelihood
confidence interval (0, 5 3 104). If our PCR conditions
caused Taq errors to occur at the estimated maximum rate,
this would only account for seven of the 58 mutations that
were present within the 13,685 bp of sequence from the
region common to all genomic cds and cDNA clones.
Furthermore, if all 58 substitutions were the result of Taq
error, then the error rate would have been 4.24 3 103
PCP Gene Family Evolution in Symbiodinium 2149
Table 3
Relative Fluorescence of DNA Stained with PicoGreen Based
on Flow Cytometry from at Least 3,000 Cells per Species
Sample
Symbiodinium 203
L. polyedra
FIG. 3.—Neighboring-Joining tree of Symbiodinium 203 genomic
and cDNA nucleotide sequences.
Parameter
Mean
CV (%)
Standard
Deviation
FL1-H
FL1-H
1.92
152.09
16.73
26.47
0.32
40.26
nium 203 genome. There was a 0.996 correlation coefficient between all points (unknowns and standards) on
the standard curve generated by quantitative real-time PCR
(curve not shown). The mean amplification value for three
replicates of 10 ng of genomic DNA was 667.60 6 51.89.
This indicates that 10 ng of Symbiodinium 203 genomic
DNA contains about the same number of PCP gene copies
as 667.60 fg of complete cds clone, which converts (with
equation 2) to 12 6 0.9 PCP genes per pg of genomic
DNA.
The genome size of Symbiodinium 203 was compared
with the 200-pg genome of L. polyedra (Holm-Hansen
1969; Spector 1984) by flow cytometry to estimate the
DNA content per nucleus. Table 3 shows the comparison
of relative mean fluorescence of PicoGreen-stained DNA
from at least 3,000 cells of each species. The flow
cytometry results indicate that Symbiodinium 203 has 3 6
1 pg DNA per genome. Although Blank and Huss (1989)
determined that the genome size of S. microadriaticum
was larger than 2 3 108 bp (0.2 pg), no other estimates for
additional species in this genus were found. By combining
the quantitative real-time PCR and flow cytometry results
within equation 1, we calculated that the Symbiodinium
203 genome contains 36 6 12 PCP gene copies.
errors/bp, considerably higher than previous estimates of
Taq error rates (First 2003).
Diversity Also Expressed at mRNA Level
The surprising diversity found in the genomic clones
was also present in cDNA clones. Poly-A mRNA from
Symbiodinium 203 was reverse transcribed with an oligo
dT primer and then cDNA amplified with the U3/L913
primer set (fig. 2d ). A single band with the expected size
was observed after gel electrophoresis, and this was
excised, purified, and cloned. As with the previous
genomic amplifications, individual cDNA clones had
diversified nucleotide sequences. Eight cDNA clones were
sequenced and none were identical to the corresponding
regions in the complete cds clones. Seven of these cDNAs
were unique. The absence of any insertions in the complete
coding sequences compared with the cDNAs is evidence
that Symbiodinium 203 PCP genes, like L. polyedra and
H. pygmaea, are intronless (Le et al. 1997; Hiller et al.
2001). The overall diversity of genomic and cDNA
nucleotide sequences is graphically depicted in the
Neighbor-Joining tree in figure 3.
Symbiodinium 203 Gene Family Size
Quantitative real-time PCR and flow cytometry were
used to estimate the number of PCP genes per Symbiodi-
Predicted Proteins, Isoelectric Points, and Amino Acid
Substitutions
There was considerable variation between amino acid
sequences predicted from complete coding regions. All
but one clone (which had a stop codon 27 bp upstream of
others) coded for 365-aa PCP preproteins with 52-aa
transit peptides and 313-aa apoproteins. Preproteins
ranged from 96.2% to 99.7% identical when compared
with each other and 93.3% to 94.9% when compared with
Symbiodinium sp.L13613. Within the Symbiodinium 203
transit peptides and apoproteins, there were nine and 20
polymorphic sites, respectively. The 313-aa apoproteins
had an average mass of 33 kDa and were the same length
as those predicted from the Symbiodinium sp.L13613
and A. carterae Z50792 and Z50793 sequences. By
contrast, the L. polyedra PCP gene sequence U93077
encodes a 375-aa preprotein with a 59-aa transit peptide
and a 316-aa apoprotein. Variability in long PCP mass is
attributable to amino acid composition and also to
polypeptide length.
Most of the predicted apoproteins from our clones
had subtle differences in their mass, but they also varied in
terms of their calculated isoelectric points (table 4). The
pIs for each of the clones fell within a range of pH 5.73 to
6.78. Most isoforms from A. carterae and Glenodinium
2150 Reichman et al.
Table 4
Calculated Mass and Isoelectric Points from Symbiodinium
203 Apoproteins
cds Clone
Number
79
80
81
82
83
84
85
87
89
90
Apoprotein Mass
(kDa)
Isoelectric Point
(pI)
33.38
31.956
33.01
32.965
33.115
33.068
32.983
32.983
33.012
32.926
6.03
6.37
6.78
6.78
5.74
6.37
6.03
6.37
5.73
5.73
(Heterocapsa) sp. have a basic pI (Haxo et al. 1976;
Prezelin and Haxo 1976). However, there are several
examples of symbiotic dinoflagellates that produce predominantly acidic PCP isoforms with pI ranges similar to
those predicted here, including S. goreauii (Trench and
Blank 1987), Symbiodinium from Montastrea annularis,
and Symbiodinium from M. cavernosa (Chang and Trench
1982). In addition, the calculated apoprotein pI of the
Symbiodinium from A. formosa sequence is 5.28.
The variability of calculated Symbiodinium 203 PCP
pIs is a direct result of predicted amino acid substitutions.
To investigate possible significance of these substitutions
to PCP function, a composite of these changes was
mapped onto the A. carterae PCP crystal structure 1PPR.
There were 70 sites at which Symbiodinium 203 cds and
A. carterae 1PPR sequences differed (excluding 59 and 39
gaps). Among these sites, 44 were fixed substitutions
between Symbiodinium 203 and A. carterae 1PPR, and 26
were polymorphic between individual Symbiodinium 203
clones. Table 5 lists the specific amino acids and types of
side chains substituted at each polymorphic site. Eleven of
the polymorphic sites (positions 6, 24,118,134,137,182,
239,244, 253, 275, and 287) were predicted to accommodate the presence or absence of amino acids with polar side
chains. Figure 4 is a rendering of all 11 polar substitutions
showing their spatial orientations in relation to the eight
peridinins within 1PPR. Thr118, Ser253, and Ser287 are
of particular interest because of their proximities to
peridinin furanic rings and polyene chains and possible
influence on spectral tuning of these chromophores (fig. 5).
The polar side chain of Thr118 is 7.19 Å from PID612.
Ser253 is 10.06 Å from PID624. Ser287 is 8.99 Å from
PID622 and 9.15 Å from PID621.
Discussion
PCP Gene family Organization and Diversity
This is the first characterization of a PCP gene family
from any dinoflagellate in the genus Symbiodinium. Earlier
descriptions of the organization of PCP genes came from
free-living species A. carterae, L. polyedra, and H.
pygmaea (Sharples et al. 1996; Le et al. 1997; Hiller
Table 5
Substitutions Within Polymorphic Sites of Predicted Symbiodinium 203 PCP Apoproteins Compared with 1PPR Amino Acids
Symbiodinium 203
Site
6
24
39
51
96
101
103
105
107
118
121
134
137
Substitutions
Ala
Thr
Asn
Asp
Glu
Lys
Gly
Glu
Ile
Val
Glu
Lys
Lys
Arg
Met
Val
Val
Ala
Lys
Thr
Ala
Ala
Asp
Val
Lys
Asn
Gln
Lys
A. carterae 1PPR
Side Chain
nonpolar
uncharged
uncharged
acidic
acidic
basic
nonpolar
acidic
nonpolar
nonpolar
acidic
basic
basic
basic
nonpolar
nonpolar
nonpolar
nonpolar
basic
uncharged
nonpolar
nonpolar
acidic
nonpolar
basic
uncharged
uncharged
basic
polar
polar
Substitutions
Side Chain
Symbiodinium 203
Site
Substitutions
Ala
nonpolar
154
Asn
uncharged polar
172
Glu
acidic
182
Ala
Arg
Ala
Val
Lys
Gln
Met
Val
Lys
Gln
Ala
Lys
Thr
Ala
Gly
Asp
Asn
Tyr
Ala
Ser
Gly
Gln
Lys
Val
Ile
Thr
Ser
Ser
Pro
Gly
nonpolar
217
Ile
nonpolar
239
Glu
acidic
Met
nonpolar
244
245
Met
nonpolar
Val
nonpolar
251
Lys
basic
253
polar
275
Ala
nonpolar
Lys
basic
Glu
acidic
276
polar
polar
279
287
NOTE.—Sites in bold accommodate the presence or absence of amino acids with polar side chains.
A. carterae 1PPR
Side Chain
nonpolar
basic
nonpolar
nonpolar
basic
uncharged
nonpolar
nonpolar
basic
uncharged
nonpolar
basic
uncharged
nonpolar
nonpolar
acidic
uncharged
uncharged
nonpolar
uncharged
nonpolar
uncharged
basic
nonpolar
nonpolar
uncharged
uncharged
uncharged
nonpolar
Substitutions
Side Chain
Ala
nonpolar
Gln
uncharged polar
Lys
basic
Ala
nonpolar
Thr
uncharged polar
Ala
nonpolar
Ala
nonpolar
Asn
uncharged polar
Ala
nonpolar
Gly
nonpolar
Val
nonpolar
Thr
uncharged polar
Ser
uncharged polar
polar
polar
polar
polar
polar
polar
polar
polar
polar
polar
PCP Gene Family Evolution in Symbiodinium 2151
FIG. 4.—Substitutions of amino acids with polar side chains
occurring at polymorphic sites within Symbiodinium 203 predicted
apoproteins shown in relation to peridinins. Numbering is based in PPR
structure.
et al. 2001). There are only two previous reports of
Symbiodinium PCP nucleotide sequences and both were
single cDNAs (Norris and Miller 1994; Weis, Verde, and
Reynolds 2002). The results of our molecular experiments
demonstrate that the PCP genes of Symbiodinium 203 are
organized essentially like those of A. carterae, L. polyedra,
and H. pygmaea in that the genes are arranged in tandem
arrays. The coding regions specify long PCP polypeptides
and are intronless. The coding regions of Symbiodinium
203 are very similar in size to those from A. carterae and
Symbiodinium from A. formosa, yet shorter than L.
polyedra. The Symbiodinium 203 PCP cassette has
untranscribed spacers of variable sizes that are smaller
than those from L. polyedra but longer than those of H.
pygmaea. The putative promoter sequence that was
identified in the same relative location upstream of
Symbiodinium 203 coding regions was 69% identical to
the L. polyedra promoter described by Li and Hastings
(1998). Dinoflagellate promoters do not fit into the
common motifs used by other eukaryotes and may be
genus or species specific.
There is far greater nucleotide diversity in the coding
regions of Symbiodinium 203 PCP genes than previously
described from any other dinoflagellate species. Eighty
nine percent of positive clones screened from genomic
and cDNA PCR libraries were distinct at the nucleotide
level. The heterogeneity of Symbiodinium 203 PCP genes
appears to be inconsistent with a pattern of concerted
evolution. Concerted evolution of tandem repeated
sequences such as the coding regions for ribosomal
RNA is usually explained in terms of two mechanisms:
continual expansion and contraction due to unequal
crossing over and biased gene conversion (see Hillis et
al. 1991). Both mechanisms rely on duplex formation
between homologous loci as most often occurs in
meiosis. Despite reported sexual reproduction in some
dinoflagellate genera and RAPDs suggesting sexual
recombination in Symbiodinium, there is an apparent
absence of a haploid sexual phase in this genus, even
FIG. 5.—Distances (E ¼ Å) between the polar side chains of (a)
threonine 118, (b) serine 253, and (c) serine 287 and the furanic ring and
polyene of nearest peridinins.
though microsatellite evidence indicates that Symbiodinium vegetative cells are haploid (Schoenberg and Trench
1980; Pfiester 1984; Trench 1993; Baillie et al. 2000;
Santos and Coffroth 2003). Homogeneity of ribosomal
coding regions within individual species suggests that
concerted evolution does occur within Symbiodinium
genomes (see Rowan and Powers 1991a,1992; Wilcox
1998). Nevertheless, the diversity of Symbiodinium 203
PCP genes is more comparable to that found in
dinoflagellate luciferin-binding protein and rubisco gene
families (Lee et al. 1993; Machabee’, Wall, and Morse
1994; Rowan et al. 1996). Perhaps mutations are
introduced into PCP coding regions faster than they can
be removed by homogenizing mechanisms. Another
possibility is that PCP genes may reside in regions of
dinoflagellate genomes where such mechanisms are less
efficient. The net effect is that PCP genes are evolving
under reduced concerted evolution.
The low Ka/Ks ratios for paired comparison between
complete coding sequences indicate that the majority of
nucleotide substitutions occur at synonymous codon
positions and that there is not a clear signal of positive
selection across entire coding regions within the Symbiodinium 203 PCP gene family. Rather, these coding regions
appear to be under purifying selection. The same is true
when the Symbiodinium 203 sequences are compared with
A. carterae or Symbiodinium from A. formosa coding
sequences (data not shown). This does not exclude the
possibility that specific codon sites within PCP genes
could be under positive selection. Analyses of PCP gene
phylogenies containing sequences from multiple species
may detect such sites.
We were concerned that at least part of the PCP gene
variation observed was the result of recombination within
the PCR reactions (Bradley and Hillis 1997). When
amplifying genes from within tandem arrays, it is possible
to generate incomplete extension products that can act as
primers and anneal to various locations in the array,
2152 Reichman et al.
resulting in PCR fragments whose sequences can be
different from those actually present in the genomic DNA.
The results of the recombination experiment showed that
8.5% of subclones had different 59 and 39 identities with
all recombinations apparently occurring in spacer regions.
Early termination of PCR extensions in the spacers may
be attributable to formation of secondary structure during
the annealing phase of the reactions. Conclusions to be
drawn from this type of recombination control are limited
in a few regards. Reamplification of small number of
cloned DNA templates can be different from amplifying
genomic regions. An increase in the number of unique
templates available within each reaction could increase
recombination frequency. Additionally, this type of
analysis only detects recombinants that are observable
by comparing coding sequence in subclones to coding
sequence in the original templates. Assuming that our
PCR conditions occasionally produced early extension
terminations as an artifact of amplifying across spacers,
it is possible that AY149122 is a chimera, but it is less
likely that the complete cds clones (AY149123 to
AY149132) or cDNA clones (AY149133 to AY149139)
are recombinants.
Another potential source of artificial substitution is
through Taq polymerase errors. Our estimated maximum
Taq error rate is 4.5 to 25 times higher than previous
reports for Taq (see references within First [2003]), yet our
maximum is still about eight times lower than the
substitution rate within the region common to genomic
cds and cDNA clones. These results strongly suggest that
the Symbiodinium 203 PCP coding sequence differences
presented here are reliable and were not caused by Taq
error.
PCP Gene Family Size
Le et al. (1997) reported that L. polyedra had a PCP
gene family of roughly 5,000 copies per 200-pg genome,
which is one of the largest gene families for any
organism. It should be noted that the L. polyedra genome
is also unusually large compared with many other
dinoflagellates and other eukaryotes (Holm-Hansen
1969; Spector 1984). For perspective, human haploid
cells typically have 3.5 pg per nucleus (Gregory 2001).
The L. polyedra copy number was based on intensity of
hybridization signals of slot blots. We sought to further
refine the methods of estimating PCP gene family size in
this investigation through a combination of quantitative
real-time PCR and flow cytometry. Quantitative real-time
PCR is an extremely accurate and reliable way of
determining the amount of starting template within
amplification reactions when compared with known
standards and was used here to give a mean number of
PCP genes per pg of genomic DNA for Symbiodinium
203. Flow cytometry is now routinely used to measure
various cellular parameters, including DNA content, and
offers the advantage of being able to quickly collect data
on large numbers of cells. Fluorescence is typically
standardized to chick red blood cells (CRBCs), chick
erythrocyte nuclei (CENs), or synthetic beads. When
Veldhuis, Cucci, and Sieracki (1997) quantified the DNA
content of 121 strains of marine phytoplankton by flow
cytometry, they pointed out that the staining of nuclei in
whole cells with cell walls was not directly comparable to
CRBCs. We chose not to standardize to CRBCs or CENs,
but rather to make a relative comparison of the mean
genome sizes of the Symbiodinium 203 and L. polyedra.
L. polyedra cells are less than perfect standards for this
purpose because they have an armored theca, whereas
Symbiodinium 203 cells do not, potentially resulting in
underestimation the genome size of Symbiodinium 203. In
addition, the wide range between the genome sizes of
these species makes calibration difficult. Nevertheless,
comparing fluorescence of stained DNA between these
two species is likely to be more relevant than comparing
either with commercially available standards. The combined results of our quantitative real-time PCR and flow
cytometry experiments show that Symbiodinium 203 has
36 6 12 PCP genes per 3 6 1 pg genome. This gene
family size is much closer to the 50 PCP genes per
genome for H. pygmaea suggested by Hiller et al. (2001)
than to L. polyedra, and the Symbiodinium 203 genome
size is comparable to A. carterae (Holm-Hansen 1969;
Spector 1984). The PCP gene copy number estimates for
L. polyedra, H. pygmaea, and Symbiodinium 203 all rely
on the accuracy of the Holm-Hansen (1969) algal DNA
content values based on amount of nuclear organic
carbon. As the DNA content of dinoflagellates is
measured with additional techniques, the current PCP
gene family sizes may be revised. If the Symbiodinium
203 PCP gene family size is accurate, then the unique
coding sequences that we identified represent 35% to
71% of the genes. Furthermore, if the diversity between
our clones reflects the overall diversity within the gene
family, then there may be as many as 42 unique PCP
coding regions present in this species.
Beyond the absolute differences in PCP gene family
sizes, Symbiodinium 203 has proportionally fewer PCP
genes per pg of genomic DNA than L. polyedra. There
may be selection in symbiotic dinoflagellates toward
smaller allocation of genomes to PCP gene families compared with free-living species. The test of this hypothesis
awaits the characterization of PCP gene families from
several more species of each type.
Affects of Genetic Diversity on Predicted PCP
Apoproteins
Although there are no reports of empirically determined pIs for Symbiodinium 203 PCP isoforms, translation of coding sequences cloned in this project
demonstrate that there is sufficient genetic diversity to
account for a suite of pIs comparable with those found in
several other Symbiodinium species through protein
analysis. This does not rule out the possibility that
posttranslational modification of PCP polypeptides could
still occur. Our results suggest that posttranslational
modifications are not necessary to explain the multiple
PCP isoforms. Isoelectric focusing of PCPs from Symbiodinium 203 could test this point, as could additional PCRbased characterizations of PCP coding sequences from
species for which PCP pIs are already known.
PCP Gene Family Evolution in Symbiodinium 2153
The potential for examining functionally significant
differences between predicted PCP apoproteins was
greatly enhanced by mapping amino acid substitutions
onto the A. carterae 1PPR crystal structure. Collectively,
the Symbiodinium 203 polypeptides differed from A.
carterae 1PPR at 70 out of 312 sites within each
monomer. The 44 fixed substitutions were distributed
through all domains of 1PPR, including regions near
adjacent monomers, the hydrophilic exterior, and chromophores. The majority of these changes that faced
hydrophobic interior of monomers did not affect the
polarity of this environment. However, they could be
important in giving Symbiodinium 203 PCP holoproteins
different conformations than 1PPR. Eleven of the 26
polymorphic sites hold amino acids with or without polar
side chains. The apoproteins from two clones contain polar
substitutions with side chain groups in positions that could
influence the spectral tuning of nearby peridinins. There
may still be additional members of the Symbiodinium
203 PCP gene family that produce isoforms with similar
effects that went undetected. Predicting the direction
and magnitude of overall changes to the spectroscopic
properties of holoproteins is beyond the scope of this
investigation. Distances calculated by between polar side
chains and peridinins could shift for individual apoproteins
if substitutions from single clones were introduced
separately within renderings.
Conclusion
Symbiodinium 203 has long PCP genes that are
intronless and arranged in tandem like those of A. carterae
and L. polyedra but with a putative promoter that is
different from L. polyedra. There are at least 17 distinct
coding regions out of 36 6 12 PCP genes in this family.
Diversity of Symbiodinium 203’s PCP gene family appears
to be consequence of low levels of concerted evolution and
provides the primary source of variability in PCP isoforms.
Amino acid substitutions in Symbiodinium 203’s PCP
apoproteins result in shifts of isoelectric points and
potentially influence light harvesting of holoproteins.
Heterogeneity in dinoflagellate PCP gene families may
provide a selective advantage as means of broadening the
range of wavelengths of light that can be captured for
photosynthesis.
Acknowledgments
We would like to thank Professor Robert K. Trench
formerly of UC Santa Barbara for donation of Symbiodinium cultures and exchange of ideas. Dr. Eric Lader
provided valuable assistance with the quantitative realtime PCR experiments. Paul Thompson, Walter Hokanson,
and Dr. E. Henry Lee assisted with statistics. Undergraduate laboratory assistants Maria Polycarpo and Tasmin
Smith gave considerable help with PCR and sequencing
routines. The Caribbean Marine Research Center Perry
Foundation, the U.S National Oceanic and Atmospheric
Administration, Oryx, Inc. and The University of Texas at
Austin provided funding for this project.
Literature Cited
Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. J.
Lipman. 1990. Basic logical alignment search tool. J. Mol.
Biol. 215:403–410.
Baillie B. K., C. A. Belda-Baillie, V. Silvestre, M. Sison, A. V.
Gomez, E. D. Gomez, and V. Monje. 2000. Genetic variation
in Symbiodinium isolates from giant clams based on randomamplified-polymorphic DNA (RAPD) patterns. Mar. Biol.
136:829–836.
Blank, R. J., and V. A. R. Huss. 1989. DNA divergency and
speciation in Symbiodinium (Dinophyceae). Plant Syst. Evol.
163:153–163.
Bradley, R. D., and D. M. Hillis. 1997. Recombinant DNA
sequences generated by PCR amplification. Mol. Biol. Evol.
14:592–593.
Chang, S. S., and R. K. Trench. 1982. Peridinin-chlorolphyll
a proteins from the symbiotic dinoflagellate Symbiodinium
(Gymnodinium) microadriaticum, Freudenthal. Proc. R. Soc.
Lond. B Biol. Sci. 215:191–210.
———. 1984. The isoelectric forms, quaternary structure and
amino acid composition of peridinin-chlorophyll a-proteins
from the symbiotic dinoflagellate Symbiodinium microadriaticum Freudenthal. Proc. R. Soc. Lond. B Biol. Sci. 222:259–271.
Coffroth, M. A., H. R. Lasker, M. E. Diamond, J. A. Bruenn, and
E. Bermingham. 1992. DNA fingerprints of a gorgonian coral:
a method for detecting clonal structure in a vegetative species.
Mar. Biol. 114:317–325.
Comeron, J. M. 1999. K-Estimator: Calculation of the number of
nucleotide substitutions per site and the confidence intervals.
Bioinformatics 15:763–764.
———. 1995. A method for estimating the numbers of
synonymous and nonsynonymous substitutions per site. J.
Mol. Evol. 41:1152–1159.
Damjanovic A., T. Ritz, and K. Schulten. 2000. Excitation
transfer in the peridinin-chlorophyll-protein of Amphidinium
carterae. Biophys. J. 79:1695–1705.
First, E. 2003. Error rates for thermal resistant DNA polymerases.
http://micro.nwfsc.noaa.gov/protocols/taq-errors.html
Govind, N. S., S. J. Roman, R. Iglesias-Prieto, R. K. Trench, E.
L. Triplett, and B. B. Prezelin. 1990. An analysis of the lightharvesting peridinin-chlorophyll a-proteins from dinoflagellates by immunoblotting techniques. Proc. R. Soc. Lond. B
Biol. Sci. 240:187–195.
Gregory, T. R. 2001. Animal genome size database. http://
www.genomesize.com.
Haxo, F. T., J. H. Kycia, G. F. Somers, A. Bennett, and H. W.
Siegelman. 1976. Peridinin-chlorolphyll a proteins of the
dinoflagellate Amphidinium carterae (Plymoth 450). Plant
Physiol. 57:297–303.
Hiller, R. G., L. G. Crossley, P. M. Wrench, N. Santucci, and
E. Hofmann. 2001. The 15-kDa forms of the apo-peridininchlorophyll a protein (PCP) in dinoflagellates show high
identity with the apo-32 kDa PCP forms, and have similar Nterminal leaders and gene arrangements. Mol. Genet. Genom.
266:254–259.
Hillis, D. M., C. Moritz, C. A. Porter, and R. J. Baker. 1991.
Evidence for biased gene conversion in concerted evolution of
ribosomal DNA. Science 251:308–310.
Hofmann, E., P. M. Wrench, F. P. Sharples, R. G. Hiller,
W. Welte, and K. Diederichs. 1996. Structural basis of light
harvesting by carotenoids: peridinin-chlorophyll-protein from
Amphidinium carterae. Science 272:1788–1791.
Holm-Hansen, O. 1969. Algae: amounts of DNA and organic
carbon in single cells. Science 163:87–88.
Iglesias-Prieto, R., N. S. Govind, and R. K. Trench. 1991. Apoprotein composition and spectroscopic characterization of the
2154 Reichman et al.
water-soluble peridinin-chlorophyll a-proteins from three symbiotic dinoflagellates. Proc. R. Soc. Lond. B Biol. Sci. 246:
275–283.
LaJeunesse, T. C. 2001. Investigating the biodiversity, ecology,
and phylogeny of endosymbiotic dinoflagellates in the genus
Symbiodinium using the ITS region: in search of a ‘‘species’’
level marker. J. Phycol. 37:866–880.
LaJeunesse, T. C., and R. K. Trench. 2000. Biogeography of two
species of Symbiodinium (Freudenthal) inhabiting the intertidal sea anemone Anthopleura elegantissima (Brandt). Biol.
Bull. 199:126–30.
Larkum, T. 1996. How dinoflagellates make light work with
peridinin. Trends Plant Sci. 1:247–248.
Le, Q. H., P. Markovic, J. W. Hastings, R. V. M. Jovine, and
D. Morse. 1997. Structure and organization of the peridininchlorophyll a-binding protein gene in Gonyaulax polyedra.
Mol. Gen. Genet. 255:595–604.
Lee, D. H., M. Mittag, S. Sczekan, D. Morse, and J. W. Hastings.
1993. Molecular cloning and genomic organization of a gene
for luciferin-binding protein from the dinoflagellate Gonyaulax polyedra. J. Biol. Chem. 268:8842–8850.
Li, L., and W. Hastings. 1998. The structure and organization of
the luciferase gene in the photosynthetic dinoflagellate
Gonyaulax polyedra. Plant Mol. Biol. 36:275–284.
Machabee’, S., L. Wall, and D. Morse. 1994. Expression and
genomic organization of a dinoflagellate gene family. Plant
Mol. Biol. 25:23–31.
Melissinos, A. C. 1966. Errors and their propagation. Pp. 467–
479 in A. C. Melissinos ed. Experiments in modern physics.
Academic Press, Orlando.
Moffat, A. S. 1996. Form follows function when plants harvest
light. Science 272:1743–1744.
Norris, B. J., and D. J. Miller. 1994. Nucleotide sequence of
a cDNA encoding the precursor of the peridinin-chlorophyll
a-binding protein from the dinoflagellate Symbiodinium sp.
Plant Mol. Biol. 24:673–677.
Ogata, T., M. Kodama, S. Nomura, M. Kobayashi, T. Nozawa, T.
Katoh and M. Mimuro. 1994. A novel peridinin-chlorophyll
a protein (PCP) from the marine dinoflagellate Alexandrium
cohorticula: a high pigment content and plural spectral forms
of peridinin and chlorophyll a. FEBS Lett. 356:367–371
Palumbi, S. R. 1996. Nucleic acids II: the polymerase chain
reaction. Pp. 205–247 in D. M. Hillis, B. K. Mable, and C.
Moritz, eds. Molecular systematics, 2nd edition. Sinauer
Associates, Sunderland, Mass.
Pfiester, L. A. 1984. Sexual reproduction. Pp. 181–199 in D. L.
Spector, ed. Dinoflagellates. Academic Press, San Diego.
Prezelin, B. B., and F. T. Haxo. 1976. Purification and
characterization of peridinin-chlorophyll a-proteins from the
marine dinoflagellates Gleodinium sp. and Gonyaulax polyedra. Planta (Berl). 128:133–141.
Rowan, R., and D. A. Powers. 1991a. A molecular genetic
classification of zooxanthellae and the evolution of animalalgal symbioses. Science. 251:1348–1351.
———. 1991b. Molecular genetic identification of symbiotic
dinoflagellates (zooxanthallae). Mar. Ecol. Prog. Ser. 71:
65–73.
———. 1992. Ribosomal RNA sequences and the diversity of
symbiotic dinoflagellates (zooxanthallae). Proc. Natl. Acad.
Sci. USA 89:3639–3643.
Rowan, R., S. M. Whitney, A. Fowler, and D. Yellowlees. 1996.
Rubisco in marine symbiotic dinoflagellates: form II enzymes
in eukaryotic oxygenic phototrophs encoded by a nuclear
multigene family. Plant Cell 8:539–553.
Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Pp. 9.31–9.37
in N. Ford, C. Nolan, and M. Ockler, eds. Molecular cloning:
a laboratory manual. Cold Spring Harbor Laboratory, Cold
Spring Harbor, N.Y.
Santos, S. R., and M. A. Coffroth. 2003. Molecular genetic
evidence that dinoflagellates belonging to the genus Symbiodinium are haploid. Biol. Bull. 204:10–20.
Schoenberg, D. A., and R. K. Trench. 1980. Genetic variation in
Symbiodinium (Gymnodinium) microadriaticum Freudenthal,
and specificity with its symbiosis with marine invertebrates. I.
Isoenzyme and soluable protein patterns of axenic cultures of
Symbiodinium microadriaticum. Proc. R. Soc. Lond. B Biol.
Sci. 207:405–427.
Sharples, F. P., P. M. Wrench, K. Ou, and R. G. Hiller. 1996.
Two distinct forms of the peridinin-chlorophyll a-protein
from Amphidinium carterae. Biochim. Biophys. Acta. 12:
1276:117–123.
Siegelman, H. W., J. H. Kycia, and F. T. Haxo. 1977. Peridininchlorophyll a-proteins of the dinoflagellate algae. Brookhaven
Symp. Biol. 28:162–169.
Song, P. S., P. Koka, B. B. Prezelin, and F. T. Haxo. 1976.
Molecular topology of the photosynthetic light-harvesting
pigment complex, peridinin-chlorophyll a-protein, from
marine dinoflagellates. Biochemistry. 15:4422–4427.
Spector, D. 1984. Dinoflagellate nuclei. Pp. 107–147 in D. L.
Spector, ed. Dinoflagellates. Academic Press, San Diego.
Swofford, D. L. 1998. PAUP*: phylogenetic analysis using
parsimony (*and other methods). Version 4. Sinauer Associates, Sunderland, Mass.
Thompson, J. D., T. J. Gibson, F. Plewniak, F. Jeanmougin, and
D. G. Higgins. 1997. The CLUSTAL_X windows interface:
flexible strategies for multiple sequence alignment aided by
quality analysis tools. Nucleic Acids Res. 25:4876–4882.
Trench, R. K. 1993. Microalgal-invertebrate symbioses: a review.
Ectocytobiosis Cell Res. 9:135–175.
Trench, R. K., and R. J. Blank. 1987. Symbiodinium microadriaticum Freudenthal, S. goreauii sp. nov., S. kawagutii sp.
nov. and S. pilosum sp. nov.: Gymnodinioid dinflagellates of
marine invertebrates. J. Phycol. 23:469–481.
Triplett, E. L., R. V. M. Jovine, N. S. Govind, S. J. Roman, S. S.
Chang, and B. B. Prezelin. 1993. Characterization of two
full-length cDNA sequences encoding for apoproteins of
peridinin-chlorophyll a-proptein (PCP) complexes. Mol.
Mar. Biol. Biotechnol. 2:246–254.
Veldhuis, M. J., T. L. Cucci, and M. E. Sieracki. 1997. Cellular
DNA content of marine phytoplankton using two new fluorochromes: taxonomic and ecological implications. J. Phycol.
33:527–541.
Weis, V. M., E. A. Verde, and W. S. Reynolds. 2002.
Characterization of a short form peridinin-chlorophyll-protein
(PCP) cDNA and protein from the symbiotic dinoflagellate
Symbiodinium muscatinei (Dinophyceae) from the sea anemone Anthopleura elegantissima (Cnidaria). J. Phycol. 38:157–
163.
Wilcox, T. P. 1998. Large-subunit ribosomal RNA systematics of
symbiotic dinoflagellates: morphology does not recapitulate
phylogeny. Mol. Phylogenet. Evol. 10:436–448.
Claudia Kappen, Associate Editor
Accepted July 30, 2003