See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/232753822

Horsetails and ferns are a monophyletic group and the closest living relatives
to seed plants

Article  in  Nature · February 2001

DOI: 10.1038/35054555

CITATIONS READS

591 1,690

7 authors, including:

Kathleen Pryer Harald Schneider

Duke University Xishuangbanna Tropical Botanical Garden
196 PUBLICATIONS   8,816 CITATIONS    356 PUBLICATIONS   11,447 CITATIONS   

SEE PROFILE SEE PROFILE

Paul G Wolf Sedonia Sipes

Utah State University Southern Illinois University Carbondale
116 PUBLICATIONS   6,343 CITATIONS    86 PUBLICATIONS   1,771 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Eupolypods II View project

Nuclear DNA C‐values in ferns and lycophytes of Argentina View project

All content following this page was uploaded by Harald Schneider on 20 May 2014.

The user has requested enhancement of the downloaded file.

letters to nature
joining trees and the amino-acid maximum parsimony phylogenies, and 100 replicates for .................................................................
Horsetails and ferns are a
the nucleotide maximum likelihood tree and the amino-acid distance-based analyses
(Dayhoff PAM matrix) (see Supplementary Information for additional trees and summary
of bootstrap support). We performed tests of alternative phylogenetic hypotheses using
Kishino±Hasegawa29 (parsimony and likelihood) and Templeton's non-parametric30 tests. monophyletic group and the
closest living relatives to seed plants
Received 30 October; accepted 4 December 2000.

1. Eisenberg, J. F. The Mammalian Radiations (Chicago Univ. Press, Chicago, 1981).

2. Novacek, M. J. Mammalian phylogeny: shaking the tree. Nature 356, 121±125 (1992).
3. O'Brien, S. J. et al. The promise of comparative genomics in mammals. Science 286, 458±481 (1999). Kathleen M. Pryer*, Harald Schneider*, Alan R. Smith²,
4. Springer, M. S. et al. Endemic African mammals shake the phylogenetic tree. Nature 388, 61±64 (1997). Raymond Cran®ll², Paul G. Wolf³, Jeffrey S. Hunt* & Sedonia D. Sipes³
5. Stanhope, M. J. et al. Highly congruent molecular support for a diverse clade of endemic African
mammals. Mol. Phylogenet. Evol. 9, 501±508 (1998).
6. McKenna, M. C. & Bell, S. K. Classi®cation of Mammals above the Species Level (Columbia Univ. Press,
* Department of Botany, The Field Museum of Natural History,
New York, 1997). 1400 S. Lake Shore Drive, Chicago, Illinois 60605, USA
7. Mouchatty, S. K., Gullberg, A., Janke, A. & Arnason, U. The phylogenetic position of the Talpidae ² University Herbarium, University of California, 1001 Valley Life Sciences
within Eutheria based on analysis of complete mitochondrial sequences. Mol. Biol. Evol. 17, 60±67 Building 2465, Berkeley, California 94720, USA
(2000).
³ Department of Biology, 5305 Old Main Hill, Utah State University, Logan,
8. Stanhope, M. J. et al. Molecular evidence for multiple origins of Insectivora and for a new order of
endemic African insectivore mammals. Proc. Natl Acad. Sci. USA 95, 9967±9972 (1998). Utah 84322, USA
9. Simpson, G. G. The principles of classi®cation and a classi®cation of mammals. Bull. Am. Mus. Nat. ..............................................................................................................................................
Hist. 85, 1±350 (1945).
10. McKenna, M. C. in Phylogeny of the Primates: a Multidisciplinary Approach (eds Luckett, W. P. &
Most of the 470-million-year history of plants on land belongs to
Szalay, F. S.) 21±46 (Plenum, New York, 1975). bryophytes, pteridophytes and gymnosperms, which eventually
11. Easteal, S. The pattern of mammalian evolution and the relative rate of molecular evolution. Genetics yielded to the ecological dominance by angiosperms 90 Myr
124, 165±173 (1990). ago1±3. Our knowledge of angiosperm phylogeny, particularly
12. Kumar, S. & Hedges, S. B. A molecular timescale for vertebrate evolution. Nature 392, 917±920
(1998).
the branching order of the earliest lineages, has recently been
13. Reyes, A. et al. Where do rodents ®t? Evidence from the complete mitochondrial genome of Sciurus increased by the concurrence of multigene sequence analyses4±6.
vulgaris. Mol. Biol. Evol. 17, 979±983 (2000). However, reconstructing relationships for all the main lineages of
14. Gregory, W. K. The orders of mammals. Bull. Am. Mus. Nat Hist. 27, 1±524 (1910).
vascular plants that diverged since the Devonian period has
15. Graur, D., Duret, L. & Gouy, M. Phylogenetic position of the order Lagomorpha (rabbits, hares and
allies). Nature 379, 333±335 (1996). remained a challenge. Here we report phylogenetic analyses of
16. Huchon, D., Catze¯is, F. M. & Douzery, E. J. P. Molecular evolution of the nuclear von Willebrand combined dataÐfrom morphology and from four genesÐfor 35
factor gene in mammals and the phylogeny of rodents. Mol. Biol. Evol. 16, 577±589 (1999). representatives from all the main lineages of land plants. We show
17. Kay, R. F., Ross, C. & Williams, B. A. Anthropoid origins. Science 275, 797±804 (1997).
18. Pettigrew, J. D. Flying Primates? Megabats have advanced pathway from eye to midbrain. Science 231,
that there are three monophyletic groups of extant vascular
1304±1306 (1986). plants: (1) lycophytes, (2) seed plants and (3) a clade including
19. Waddell, P. J., Okada, N. & Hasegawa, M. Towards resolving the interordinal relationships of placental equisetophytes (horsetails), psilotophytes (whisk ferns) and all
mammals. Syst. Biol. 48, 1±5 (1999). eusporangiate and leptosporangiate ferns. Our maximum-like-
20. Krettek, A., Gullberg, A. & Arnason, U. Sequence analysis of the complete mitochondrial DNA
molecule of the hedgehog, Erinaceus europaeus, and the phylogenetic position of the Lipotyphla.
lihood analysis shows unambiguously that horsetails and ferns
J. Mol. Evol. 41, 952±957 (1995). together are the closest relatives to seed plants. This refutes the
21. Rose, K. D. & Emry, R. J. in Mammal Phylogeny: Placentals (eds Szalay, F. S., Novacek, M. J. & prevailing view that horsetails and ferns are transitional evolu-
McKenna, M. C.) 81±102 (Springer, New York, 1993).
tionary grades between bryophytes and seed plants7, and has
22. Miyamoto, M. M. & Goodman, M. Biomolecular systematics of eutherian mammals: phylogenetic
patterns and classi®cation. Syst. Zool. 35, 230±240 (1986).
important implications for our understanding of the development
23. Madsen, O. et al. Parallel adaptive radiations in two major clades of placental mammals. Nature (in the and evolution of plants8.
press). Estimates of a phylogeny for the main groups of land plants, each
24. Eizirik, E., Murphy, W. J. & O'Brien, S. J. Molecular dating and biogeography of the early placental
mammal radiation. J. Hered. (in the press).
with highly divergent morphologies, have been many, and all have
25. Hedges, S. B., Parker, P. H., Sibley, C. G. & Kumar, S. Continental breakup and the ordinal been contested. Bryophytes (liverworts, hornworts and mosses) are
diversi®cation of birds and mammals. Nature 381, 226±229 (1996). consistently shown to be a basal grade, but their relationships to one
26. Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F. & Higgins, D. G. The Clustal X windows another and to vascular plants are controversial1,2,9±13. Most phylo-
interface: ¯exible strategies for multiple sequence alignment aided by quality analysis tools. Nucl.
Acids Res. 25, 4876±4882 (1997).
genetic analyses of vascular plants consistently reconstruct two
27. Swofford, D. L. PAUP* Phylogenetic Analysis using Parsimony and Other Methods (Sinauer, Sunder- main lines of evolution: the Lycophytina (clubmosses and relatives),
land, Massachusetts, 1998). with 1% of extant diversity, and the Euphyllophytina (all other
28. Strimmer, K. & Von Haeseler, A. Quartet puzzling: a quartet maximum likelihood method for vascular plants)1,2,10,11,14±17. Extant Euphyllophytina1,2 comprises six
reconstructing tree topologies. Mol. Biol. Evol. 13, 964±969 (1996).
29. Kishino, H. & Hasegawa, M. Evaluation of the maximum likelihood estimate of the evolutionary tree major monophyletic lineages: Equisetopsida (horsetails), Polypo-
topologies from DNA sequence data, and the branching order in Hominoidea. J. Mol. Evol. 29, 170± diidae (leptosporangiate ferns), Spermatophytata (seed plants),
179 (1989). Psilotidae (whisk ferns; simple plants regarded by some to be
30. Templeton, A. R. Phylogenetic inference from restriction endonuclease cleavage site maps with
particular reference to the evolution of humans and the apes. Evolution 37, 221±244 (1983).
living relicts of the earliest vascular plants7,18), Marattiidae and
Ophioglossidae (eusporangiate ferns). Phylogenetic assessments
Supplementary information is available on Nature's World-Wide Web site based on single genes10,11,14,15,19 and/or morphology1,7,12,17,20 have
(http://www.nature.com) or as paper copy from the London editorial of®ce of Nature.
Sequences are deposited in GenBank under accession numbers AY011125±AY012154.
provided only weak and usually contradictory evidence for the
relationships among these euphyllophyte lineages. Resolving these
relationships would not only stabilize a pivotal region of vascular
Acknowledgements
plant phylogeny but is also key to identifying the most appropriate
We thank D. Hirschmann, M. Houck, R. Montali, R. Baker, G. Harris, K. Helgen,
A. L. Roca, M. Roelke-Parker, A. Grafodatsky, O. Serov and T. Oleksyk for help in
outgroup for addressing questions related to the evolution of seed
obtaining samples and technical assistance, and M. Smith and M. Dean for helpful plants.
suggestions. We also thank the NCI Frederick Molecular Technology Center for technical Recent palaeontological studies1,2,7 attempted to demonstrate that
support, and the Advanced Biomedical Computer Center for computational assistance. approaches based solely on living species would have dif®culties
All tissue samples were obtained with appropriate permits (CITES) issued to the National
Cancer Institute, National Institutes of Health (principal of®cer, S.J.O). Y.P.Z. is supported
reconstructing relationships among major lineages of vascular
by the Natural Science Foundation of China and Chinese Academy of Sciences. E.E. is plants. Inadequate taxon sampling, rate heterogeneity across DNA
supported by Conselho Nacional de Desenvolvimento Cientõ®co e TecnoloÂgõÂco (CNPq), nucleotide sites among lineages, and inappropriate algorithms also
Brazil. have been cited as impediments to resolving ancient branching
Correspondence and requests for materials should be addressed to S.J.O. events21. However, as predicted by recent theoretical studies22,
(e-mail: [email protected]). combined analysis of DNA sequences from multiple loci proves to

618 © 2001 Macmillan Magazines Ltd NATURE | VOL 409 | 1 FEBRUARY 2001 | www.nature.com
 letters to nature
be very useful in inferring deep phylogenetic patterns4±6. With few We obtained DNA sequences (5,072 aligned base pairs) of four
exceptions12,20, broad phylogenetic studies rely solely on combined genes from two plant genomes: plastid atpB, rbcL and rps4, and
nucleotide sequence data, with authors arguing that morphological nuclear small-subunit ribosomal DNA. We also assembled a con-
character homology assessment among ancient and divergent gruent data set of 136 vegetative and reproductive morphological/
groups is too challenging. This practice ignores the higher complex- anatomical characters. We sampled 35 representatives from all major
ity of morphological characters that can conserve character states monophyletic lineages of land plants. The selection of taxa re¯ects
over time and that have a lower probability of random evolution of our focus on basal vascular plants, and all six Euphyllophytina1
similar structures. lineages are represented by two or more members. Five bryophytes

100/100 Pteridium
Blechnum
90/91 polypodiaceous
ferns
-/- Dicksonia
Plagiogyria
100/100 86/100 Cyathea

leptosporangiate ferns
Polypodiidae
tree ferns
88/96 100/100 Marsilea
Salvinia
-/64 Lygodium heterosporous
water ferns

Moniliformopses
100/100 Phanerosorus
70/81 Gleichenia
100/100 Hymenophyllum gleichenioid ferns

Osmunda

Euphyllophytina
87/56

100/100 Osmunda Equisetum 1

Equisetum 2
62/- Angiopteris
92/100 Equisetopsida
100/98
100/100 Marattia horsetails
Danaea
Marattiidae
eusporangiate
100/100 Ophioglossum ferns
Botrychium
100/100 Ophioglossidae
92/-
100/100 Psilotum
Tmesipteris

Psilotidae
whisk ferns
89/- Gnetum
Pinus
Spermatophytata

65/-

100/96 Cycas
100/100 55/- Ginkgo

100/100 Austrobaileya
Chloranthus
Spermatophytata
seed plants
Lycophytina

75/93 Selaginella
100/99 Isoetes
Huperzia
Lycophytina
Bryophytes

91/- Polytrichum
69/- Sphagnum
94/59 Marchantia
Haplomitrium
Anthoceros bryophytes

0.1 substitutions per site

Figure 1 Phylogenetic relationships for all the main lineages of vascular plants inferred them are arbitrary. Branches leading to the three monophyletic clades of vascular plants
from maximum-likelihood (ML) analysis of the combined chloroplast rbcL, atpB, rps4 and (lycophytes, seed plants and horsetails+ferns) are drawn medium thick. The branch
nuclear small-subunit rDNA data set. Numbers at nodes and before the slash are ML supporting the Euphyllophytina, with horsetails+ferns as sister group to seed plants, is the
bootstrap values $50%; maximum parsimony (MP) bootstrap values $50% appear after thickest. Wiggled lines (at straight arrows) indicate three areas of con¯ict between the ML
the slash when these same nodes were supported in the MP unequally weighted analysis and MP analyses. Branch lengths are proportional to number of substitutions per site
of the combined four-genes plus morphology data set (single MP tree = 14165.04 steps). (scale bar). Thumbnail sketches of plant representatives accompany major clades.
A minus sign indicates a node had less than 50% bootstrap support in one or the other Taxonomy follows ref. 1.
analysis. The topology is rooted by all bryophytes, hence relationships depicted among

NATURE | VOL 409 | 1 FEBRUARY 2001 | www.nature.com © 2001 Macmillan Magazines Ltd 619
letters to nature
were speci®ed as outgroups. We analysed the data sets using our knowledge, this relationship has been proposed only once
both maximum-parsimony (MP) and maximum-likelihood (ML) previously1, as a tentative hypothesis on the evidence of a single
optimization criteria; bootstrap (BS) analyses were conducted to anatomical character (protoxylem distribution). This led to the
measure the stability of observed phylogenetic patterns. provisional classi®cation of the horsetail±fern clade as infradivision
Using ML on the combined four-gene data set we recovered one Moniliformopses (moniliforms); Psilotidae, however, was not
most likely tree (-ln likelihood = 36466.6365) for each of the 100 included in that study1. Although this same deep dichotomy is
replicates (Fig. 1). We also observed an essentially identical topology also robustly resolved in the MP analysis of the combined four-
using MP on the combined four-gene and morphology data set genes plus morphology data set, the Euphyllophytina node is weakly
(three areas that differ are highlighted on Fig. 1). Regardless of the supported (,50% BS). Exceptionally long branches in each of the
analytical approach (MP or ML), three major lineages emerged as three main clades (Fig. 1: Selaginella, Gnetum and Equisetum) and
monophyletic clades with exceptional support (100% BS). The ®rst the greater sensitivity of MP over ML to long-branch attraction
clade comprises the Lycophytina, increasingly recognized as a (statistical inconsistency) effects21,25 probably explain why par-
distinct group of vascular plants only distantly related to other simony bootstrapping failed to recover this clade with high con-
extant pteridophytes and seed plants1,16. The second diverging ®dence. When these long-branch taxa were removed and the
lineage corresponds to seed plants. The third, novel, clade includes combined four-genes plus morphology data set was re-analysed
all non-seed-producing lineages of Euphyllophytina, including with MP, this same basal Euphyllophytina node was highly sup-
horsetails (Equisetopsida), leptosporangiate ferns (Polypodiidae), ported (83% BS, results not shown). Each of our separate single-
eusporangiate ferns (Marattiidae, Ophioglossidae) and whisk ferns gene analyses, with the exception of rps4, did not resolve the
(Psilotidae). Seed plants, ferns and horsetails are united as a horsetail±fern clade, and none was able to determine con®dently
monophyletic group, to the exclusion of lycopods, in both the ML the closest relatives to seed plants. Only our morphological data set,
(92% BS) and MP (,50% BS) analyses. when analysed alone with MP, provided the same conclusions
We observed one unambiguous length discrepancy in rps4 that
can be interpreted as a molecular `signature' providing additional
support for horsetail±fern monophyly. A portion of the rps4
alignment is shown for base pairs 646±696 (Fig. 2), which includes
27 ambiguously aligned base pairs (658±684) ¯anked by unam-
biguously aligned regions. The ambiguously aligned region was
excluded entirely from the ML analysis. In the MP analysis, the same
region was recoded simply as a single absence/presence character for
the observed length increase. This multi-residue length increase in
horsetails and ferns is not as likely to be a random convergence as is
a single point mutation and provides further evidence for this clade.
Within the horsetail±fern lineage, Psilotidae is most closely
related to Ophioglossidae (100% BS). Although this association
was only weakly suggested in recent single-gene analyses11,19,20, the
current evidence unambiguously invalidates the traditional mor-
phological and palaeobotanical view that Psilotidae are relatively
unaltered descendants of the psilotophytes, among the earliest
vascular plant fossils7,18. Ophioglossidae and Psilotidae differ so
radically in phenotype that this close relationship, implying a shared
origin of phenotypic simpli®cation, was never before explicitly
considered. All other ferns and horsetails make up its sister clade
(87% BS). The relationships of horsetails also have been con-
troversial: sister to seed plants7, sister to leptosporangiate (Poly-
podiidae) and eusporangiate (Ophioglossidae and Marattiidae)
ferns1, or as a basal grade euphyllophyte lineage17. Our analysis
clearly (100% BS) places Equisetum within the non-lycophyte
pteridophyte clade, although its exact relationships within this
clade are not yet well resolved. In the ML analysis, Equisetum is
sister to Marattiidae (62% BS), whereas in the MP analysis, it is
sister to leptosporangiate ferns (,50% BS). This study also con-
®rms a sister relationship between tree ferns and the more derived
`polypodiaceous' leptosporangiate ferns (90% BS), and places the
heterosporous water ferns as sister to this clade (100% BS) (Fig. 1).
Relationships among these groups were equivocal in earlier
studies17,20.
The only noteworthy disagreement between our ML and MP
analyses is localized within seed-plant relationships, a subject of
much current controversy21,23,24. Our ML analysis resolved gymno-
sperms as monophyletic (65% BS) and Gnetum as sister to Pinus
(89% BS). Our MP analysis supports Gnetum as basal among seed
plants (87% BS), and all other gymnosperms as monophyletic (67%
BS) and sister to angiosperms.
In the ML analysis of the combined four-gene data set, there is Figure 2 A portion of the chloroplast rps4 alignment. An ambiguously aligned region (grey
persuasive support for the Euphyllophytina (92% BS), with a basal box) containing a 9-base-pair length difference distinguishes horsetails and ferns (bottom
dichotomy indicating that the horsetail±fern clade (100% BS) is block) from bryophytes, lycophytes and seed plants (top block). Amino-acid translations
the closest relative to seed plants (100% BS). To the best of are interleaved below each DNA sequence. Dashes indicate gaps.

620 © 2001 Macmillan Magazines Ltd NATURE | VOL 409 | 1 FEBRUARY 2001 | www.nature.com
 letters to nature
regarding the Euphyllophytina as when the four genes were analysed probabilities for each of six substitution types, plus three heterogeneous rate categories
across sites following a discrete approximation of the gamma distribution, 100 random-
simultaneously with ML. A study using mitochondrial small-sub-
addition replicates) analyses using PAUP* version 4.0b2a30. The ML analysis was restricted
unit rDNA sequence data10 with a smaller selection of taxa suggested to the combined four-gene data set because it is not possible to simultaneously implement
support for this hypothesis; however, critical euphyllophyte taxa two models of evolution, one for morphology and one for DNA sequence data, in any
(Psilotidae and Marattiidae) were not included. A more recent currently available computer programs. We further performed both parsimony bootstrap
study26 that combines data from two genes (nuclear and mitochon- (unequal weighting schemes, 1,000 replicates, each with 10 random-addition replicates
and TBR branch swapping) and likelihood bootstrap analyses (212 replicates, using
drial small-subunit rDNA) strongly corroborates a horsetail±fern identical parameters to those used to ®nd the most likely tree).
clade as sister to seed plants, despite a limited sampling of only seven
Received 25 July; accepted 27 November 2000.
euphyllophyte taxa from all pertinent lineages.
Our report of a basal dichotomy in the Euphyllophytina, a split 1. Kenrick, P. & Crane, P. R. The Origin and Early Diversi®cation of Land Plants: A Cladistic Study
(Smithsonian Institution Press, Washington DC, 1997).
that occurred in the early±mid Devonian (about 400 Myr ago),
2. Kenrick, P. & Crane, P. R. The origin and early evolution of plants on land. Nature 389, 33±39 (1997).
necessitates abandonment of the prevailing view that ferns and 3. Lupia, R., Lidgard, S. & Crane, P. R. Comparing palynological abundance and diversity: Implications
horsetails represent paraphyletic successive grades of increasing for biotic replacement during the Cretaceous angiosperm radiation. Paleobiology 25, 305±340 (1999).
complexity in early vascular plant evolution, which eventually led 4. Soltis, P. S., Soltis, D. E. & Chase, M. W. Angiosperm phylogeny inferred from multiple genes as a tool
for comparative biology. Nature 402, 402±404 (1999).
to the more complex seed plants, and ultimately to angiosperms. 5. Qiu, Y.-L. et al. The earliest angiosperms: Evidence from mitochondrial, plastid and nuclear genomes.
A parallel deep reorganization of metazoan phylogeny has recently Nature 402, 404±407 (1999).
been proposed27, with `simple' bilaterian taxa (for example, platy- 6. Barkman, T. J. et al. Independent and combined analyses of sequences from all three genomic
helminths and nemerteans) being displaced from the base of the compartments converge on the root of ¯owering plant phylogeny. Proc. Natl Acad. Sci. USA 97,
13166±13171 (2000).
metazoan tree to within the large lophotrochozoan clade. 7. Rothwell, G. W. Fossils and ferns in the resolution of land plant phylogeny. Bot. Rev. 65, 188±218
A corollary of the demise of the paraphyletic interpretation of (1999).
early vascular plant evolution is that it is now necessary to confront 8. Graham, L. E., Cook, M. E. & Busse, J. S. The origin of plants: Body plan changes contributing to a
major evolutionary radiation. Proc. Natl Acad. Sci. USA 97, 4535±4540 (2000).
the many recurring models that derive morphological, develop-
9. Bremer, K. Summary of green plant phylogeny and classi®cation. Cladistics 1, 369±385 (1985).
mental and physiological conditions in seed plants from an `inter- 10. Duff, R. J. & Nickrent, D. L. Phylogenetic relationships of land plants using mitochondrial small-
mediate' or `primitive' pteridophyte ancestor. We predict that this subunit rDNA sequences. Am. J. Bot. 86, 372±386 (1999).
will require a signi®cant revision in the interpretation of the 11. Hedderson, T. A., Chapman, R. & Cox, C. J. in Bryology for the Twenty-®rst Century (eds Bates, J. W.,
Ashton, N. W. & Duckett, J. G.) 65±77 (Maney Publishing and the British Bryological Society, Leeds,
underlying processes of vascular plant evolution. For example, a 1998).
number of homeotic genes, such as the MADS-box genes that 12. Mishler, B. D. et al. Phylogenetic relationships of the ``green algae'' and ``bryophytes''. Ann. Missouri
encode transcription factors critical for regulating physiological Bot. Gard. 81, 451±483 (1994).
and developmental processes, especially ¯ower development, have 13. Qiu, Y.-L., Cho, Y., Cox, J. C. & Palmer, J. D. The gain of three mitochondrial introns identi®es
liverworts as the earliest land plants. Nature 394, 671±674 (1998).
been well studied in angiosperms28. Clarifying the origin of these 14. Hiesel, R., von Haeseler, A. & Brennicke, A. Plant mitochondrial nucleic acid sequences as a tool for
genes has been hampered by the few reports of homologues from phylogenetic analysis. Proc. Natl Acad. Sci. USA 91, 634±638 (1994).
non-seed plants, and therefore it is not known to what extent 15. Kranz, H. D. & Huss, V. A. R. Molecular evolution of pteridophytes and their relationships to seed
plants: Evidence from complete 18S rRNA gene sequences. Plant Syst. Evol. 202, 1±11 (1996).
changes in number, regulation and function of these and other
16. Raubeson, L. A. & Jansen, R. K. Chloroplast DNA evidence on the ancient evolutionary split in
homeotic genes may have driven land plant evolution. The study of vascular land plants. Science 255, 1697±1699 (1992).
these genes from across a stable phylogenetic framework is critical. 17. Stevenson, D. W. & Loconte, H. in Pteridology in Perspective (eds Camus, J. M., Gibby, M. & Johns,
We note that all the main plant model organisms (for example, R. J.) 435±467 (Royal Botanic Gardens, Kew, 1996).
18. Wagner, W. H. Jr Systematic implications of the Psilotaceae. Brittonia 29, 54±63 (1977).
Arabidopsis, Glycine, Lycopersicon, Oryza, Petunia and Zea) are 19. Manhart, J. R. Phylogenetic analysis of green plant rbcL sequences. Mol. Phylogenet. Evol. 3, 114±127
recently evolved angiosperms. Efforts to promote developmental (1994).
and genomic research on model systems in the horsetail±fern clade 20. Pryer, K. M., Smith, A. R. & Skog, J. E. Phylogenetic relationships of extant ferns based on evidence
(for example, Ceratopteris29), will probably lead to an improved from morphology and rbcL sequences. Am. Fern J. 85, 205±282 (1995).
21. Sanderson, M. J., Wojciechowski, M. F., Hu, J. M., Sher Khan, T. & Brady, S. G. Error, bias, and long-
understanding of fundamental aspects of vascular plant develop- branch attraction in data for two chloroplast photosystem genes in seed plants. Mol. Biol. Evol. 17,
ment and evolution8. M 782±797 (2000).
22. Hillis, D. M. Inferring complex phylogenies. Nature 383, 130±131 (1996).
23. Bowe, L. M., Coat, G. & dePamphilis, C. W. Phylogeny of seed plants based on all three genomic
Methods compartments: Extant gymnosperms are monophyletic and Gnetales' closest relatives are conifers.
Taxon sampling and morphological data set Proc. Natl Acad. Sci. USA 97, 4092±4097 (2000).
24. Chaw, S.-M., Parkinson, C. L., Cheng, Y., Vincent, T. M. & Palmer, J. D. Seed plant phylogeny inferred
We selected 35 taxa to sample explicitly at least two members of each major monophyletic from all three plant genomes: Monophyly of extant gymnosperms and origin of Gnetales from
group of land plants. The various groups were determined from recent broad-scale conifers. Proc. Natl Acad. Sci. USA 97, 4086±4091 (2000).
phylogenetic analyses1,12,17,20, and we speci®ed the bryophytes Anthoceros, Haplomitrium, 25. Lewis, P. O. in Molecular Systematics of Plants II; DNA Sequencing (eds Soltis, D. E., Soltis, P. S. &
Marchantia, Polytrichum and Sphagnum as outgroups. Our morphological data set Doyle, J. J.) 132±163 (Kluwer Academic, Boston, 1998).
comprised 136 parsimony-informative characters (H.S. et al., manuscript in preparation), 26. Renzaglia, K. S., Duff, R. J., Nickrent, D. L. & Garbary, D. J. Vegetative and reproductive innovations of
which we, for the most part, adopted or modifed from recent studies1,7,12,17,20. early land plants: Implications for a uni®ed phylogeny. Phil. Trans. R. Soc. Lond. B 355, 769±793
(2000).
Gene sequencing 27. Adoutte, A. et al. The new animal phylogeny: Reliability and implications. Proc. Natl Acad. Sci. USA
97, 4453±4456 (2000).
We ampli®ed chloroplast rbcL, atpB, rps4, and nuclear small-subunit rDNA genes for all 35
28. Theissen, G. et al. A short history of MADS-box genes in plants. Plant Mol. Biol. 42, 115±149 (2000).
taxa from total cellular DNA by polymerase chain reaction (PCR) and sequenced them
29. Hickok, L. G., Warne, T. R. & Fribourg, R. S. The biology of the fern Ceratopteris and its use as a model
using an ABI 377 automated DNA sequencer (PE Applied Biosystems). Details of taxon
system. Int. J. Plant Sci. 156, 332±345 (1995).
sampling, DNA isolation, PCR ampli®cation, sequencing, sequence alignment, exclusion 30. Swofford, D. L. PAUP*. Phylogenetic Analysis Using Parsimony (*and other Methods) (Sinauer,
and recoding of ambiguously aligned regions, data set combinability testing, and Sunderland, Massachusetts, 1999).
phylogenetic analysis will be published elsewhere (K.M.P. et al., manuscript in prepara-
tion). Most atpB, rps4, nuclear small-subunit rDNA, and some rbcL sequences were Supplementary information is available on Nature's World-Wide Web site
generated as part of this study. For voucher information, GenBank numbers and the (http://www.nature.com) or as paper copy from the London editorial of®ce of Nature.
aligned data matrices, see Supplementary Information and http://www.fmnh.org/
research_collections/botany/botany_sites/ferns/publications.html; data matrices are also
Acknowledgements
available in TreeBASE, accession number S543, at
http://www.herbaria.harvard.edu/treebase/. We thank R. Lupia, F. M. Lutzoni, B. D. Mishler, L. Newstrom-Lloyd and S. Zoller for
critical comments on the manuscript; Z. Dabich, J. BeÂlisle, R. Lupia and D. Kieffer for
assistance in rendering Fig. 1; F. M. Lutzoni and V. A. Funk for advice on phylogenetic
Phylogenetic analyses analyses; I. Capesius, S. Boyles, B. Gof®net, M. Hasebe, M. Kato, M. Kessler, B. D. Mishler,
We conducted heuristic MP (unequal weighting schemes, 1,000 random-addition R. Moran, J. Shaw, W. C. Taylor, Y.-L. Qiu, D. Wall, J. Wheeler, and greenhouse managers at
replicates, tree bisection-reconnection (TBR) branch swapping) and ML (general time- Humboldt State University, University of California at Davis, University of California
reversible model, accommodating unequal nucleotide frequencies and different- Botanical Garden at Berkeley, New York Botanical Garden for plant material; S. W.

NATURE | VOL 409 | 1 FEBRUARY 2001 | www.nature.com © 2001 Macmillan Magazines Ltd 621
letters to nature
Graham, P. S. Soltis and J. Therrien for sharing unpublished sequence data; and examined the interaction of SNO±Hb5,6 with inside-out vesicles
D. Ferguson, E. Grismer, J. Irwin and L. Sappelsa for general assistance in the initial stages
(IOVs) prepared by everting RBC membrane ghosts11. IOVs incu-
of the project. This work was supported by grants from the NSF to K.M.P., A.R.S., P.G.W.
and R. C., the Green Plant Phylogeny Research Coordination Group (USDA grant), and by bated with SNO±Hb and washed at pH 8 to remove bound Hb
the Pritzker Foundation Fund of The Field Museum. incorporated about 450 pmol NO per mg of TX100-extracted IOV
protein (Fig. 1d). All the incorporated NO was present in complex
Correspondence and requests for materials should be addressed to K.M.P.
(e-mail: kpryer@®eldmuseum.org). with thiol, that is, as SNO. It is important to note that SNO was not
detected in extracts of IOVs exposed to NO in the absence of Hb
(data not shown).
To rule out the possibility that apparent NO group transfer
to IOVs was an artefact of residual membrane-bound SNO±Hb,
................................................................. we incubated IOVs with SNO±Hb immobilized on Sephadex
beads. After centrifugal separation, washes at pH 7 and solubiliza-
Export by red blood cells tion in TX100, extracts of IOVs were free of Hb as assessed by
spectrophotometric detection of haem. SNO was present in those
of nitric oxide bioactivity extracts at somewhat higher levels than in extracts derived from
IOVs incubated with free SNO±Hb (suggesting a greater loss of
John R. Pawloski, Douglas T. Hess & Jonathan S. Stamler

Howard Hughes Medical Institute and Department of Medicine, Box 2612, a cytosol
Duke University Medical Center, Durham, North Carolina 27710, USA 120
81 membrane
75
.............................................................................................................................................. 100
65

%NO recovered
59
Previous studies support a model in which the physiological O2 80

gradient is transduced by haemoglobin into the coordinate release 60

from red blood cells of O2 and nitric oxide (NO)-derived vasoac- 40

22 23
20
tivity to optimize oxygen delivery in the arterial periphery1,2. But 20
17

whereas both O2 and NO diffuse into red blood cells, only O2 can 0
diffuse out3±5. Thus, for the dilation of blood vessels by red blood 1:1000 1:500 1:250 1:100

cells, there must be a mechanism to export NO-related vasoactiv- b

120
ity, and current models of NO-mediated intercellular commu-
nication should be revised. Here we show that in human 100

erythrocytes haemoglobin-derived S-nitrosothiol (SNO), gener-

% Fe[ II ] NO

80
ated from imported NO, is associated predominantly with the red 60
blood cell membrane, and principally with cysteine residues in the 40
haemoglobin-binding cytoplasmic domain of the anion exchan- 20
ger AE1. Interaction with AE1 promotes the deoxygenated struc- 0
ture in SNO±haemoglobin, which subserves NO group transfer to 1:1000 1:500 1:250 1:100
the membrane. Furthermore, we show that vasodilatory activity is c
released from this membrane precinct by deoxygenation. Thus, 120

the oxygen-regulated cellular mechanism that couples the synthe- 100

sis and export of haemoglobin-derived NO bioactivity operates, at 80

%SNO

least in part, through formation of AE1±SNO at the membrane± 60

cytosol interface. 40
As the ®rst step in analysing the fate of haemoglobin (Hb)- 20
derived NO in situ, we determined the disposition of NO groups
0
transfered physiologically from the haems of Hb to b-chain Cys 93 1:1,000 1:500 1:250 1:100
in intact human erythrocytes3,4. Red blood cells (RBCs) held at less NO:haem
than 1% O2 were exposed for 5 min to physiological amounts of NO d
(100 nM to 1 mM; NO:haem ratios 1:1,000 to 1:100) followed by 900
reoxygenation (21% O2), and membrane and cytosolic fractions 800
SNO (pmol mg–1 protein)

were prepared. Fractions were solubilized with Triton X-100 700

(TX100), and the NO content of extracts was measured by photo- 600
lysis/chemiluminescence3,4. At the lower NO:haem ratios, which 500 SNO-Hb
sepharose
produced intracellular NO concentrations matching those found in 400 SNO-Hb
vivo (100±800 nM), recovery of NO was essentially complete, that 300
free

is, none was lost to nitrate (Fig. 1a). In this model system, about 15± 200
20% of NO incorporated by RBCs was present as SNO; the 100
remainder was ascribed largely to iron nitrosyl haem (FeNO)1,3,4,6. PCMPS Chymo

Most iron nitrosyl Hb was recovered with the cytosolic fraction

(Fig. 1b). In contrast, SNO was associated predominantly with the Figure 1 Haemoglobin-derived SNO is associated with cysteine thiols of RBC membrane
membrane fraction (Fig. 1c). These results con®rm that, in intact proteins. a±c, Distribution in cytosolic and membrane fractions of NO groups after
RBCs7 as with isolated reactants3,4, Hb will ef®ciently capture and exposure of intact RBCs to NO. Recovery of NO is essentially complete at low,
preserve NO, and form SNO, under physiological conditions. physiological NO:haem ratios (a), which yield 100±800 nM intracellular NO; FeNO is
Unexpectedly, however, the formation of SNO is compartmenta- predominantly cytosolic (b), whereas SNO is largely membrane associated (c) (P , 0.05
lized within the RBC. for all pairwise comparisons). d, SNO content of IOVs exposed to free or Sepharose-bound
Haemoglobin associates with the cytoplasmic face of the RBC SNO±Hb (50 nmol SNO±Hb per mg IOV protein). Transfer of NO groups to the membrane
membrane through speci®c protein±protein interactions8±10. To is greatly reduced (P , 0.05) after treatment of IOVs with the thiol-modifying reagent
determine the disposition of Hb-derived membrane SNO, we PCMPS and after mild digestion of IOVs with chymotrypsin (chymo). (n = 3±7 for a±d.)

622 © 2001 Macmillan Magazines Ltd NATURE | VOL 409 | 1 FEBRUARY 2001 | www.nature.com
View publication stats