Tracheid Structure in a Primitive Extant Plant Provides an Evolutionary Link to Earliest Fossil Tracheids

TRACHEID STRUCTURE IN A PRIMITIVE EXTANT PLANT PROVIDES AN EVOLUTIONARY LINK TO EARLIEST FOSSIL TRACHEIDS

Martha E. Cook and William E. Friedman1

Department of Environmental, Population, and Organismic Biology, University of Colorado, Boulder, Colorado 80309, U.S.A.

Most attempts to understand the early evolution of tracheids have centered on fossil Silurian and Devonian vascular plants, and these efforts have led to a wealth of new information on early water-conducting cells. All of these early tracheids appear to possess secondary cell wall thickenings composed of two distinct layers: a layer adjacent to the primary cell wall that is prone to degradation (presumably during the process of fossilization) and a degradation-resistant (possibly lignified) layer next to the cell lumen. Developmental studies of secondary wall formation in tracheary elements of extant vascular plants have been confined to highly derived seed plants, and it is evident that the basic structure of these secondary cell wall thickenings does not correspond well to those of tracheids of the Late Silurian and Early Devonian. Significantly, secondary cell wall thickenings of tracheary elements of seed plants are not known to display the coupled degradation-prone and degradation-resistant layers characteristic of tracheids in early tracheophytes. We report a previously unknown pattern of cell wall formation in the tracheids of a living plant. We show that in Huperzia, one of the most primitive extant vascular plants, secondary cell wall deposition in tracheids includes a first-formed layer of wall material that is degradation-prone (“template layer”) and a later-formed degradation-resistant layer (“resistant layer”). These layers match precisely the pattern of wall thickenings in the tracheids of early fossil vascular plants and provide an evolutionary link between tracheids of living vascular plants and those of their earliest fossil ancestors. Moreover, our developmental data provide the essential information for an explicit model of the early evolution of tracheid secondary wall thickenings. Finally, congruence of tracheid structure in extant Huperzia and Late Silurian and Early Devonian vascular plants supports the hypothesis of a single origin of tracheids in land plants.

Introduction Edwards 1993). Recent phylogenetic analyses indicate that tracheid-bearing plants (tracheophytes) are monophyletic (Ken- The early evolution of vascular plants (tracheophytes) in the rick and Crane 1991, 1997a) and that diversification among Silurian constitutes the first major diversification of photosyn- early tracheophytes produced three major clades (Banks 1975; thetic life on land (Kenrick and Crane 1997a, 1997b). While Kenrick and Crane 1991, 1997a; fig. 1). The earliest members there is evidence for the establishment of terrestrial plant life of each clade are characterized by a distinctive tracheid type by the end of the Ordovician (Gray et al. 1982; Gray 1993), (Kenrick and Crane 1991, 1997a). the fossil record indicates that land plants remained extremely Rhyniopsida is hypothesized to be an early divergent mon- small and structurally simple until the Late Silurian (Knoll and ophyletic clade (or possibly paraphyletic grade) of primitive Rothwell 1981; Gensel and Andrews 1987; Knoll and Niklas vascular plants that is sister to a monophyletic eutracheophyte 1987; Kenrick and Crane 1997a, 1997b). Among the events clade that includes all extant vascular plants as well as many thought to have been associated with the first burst of struc- of their extinct relatives (Kenrick and Crane 1991, 1997a, and tural diversification among land plants is the evolution of tra- references therein). Members of the Rhyniopsida (all extinct) cheids, complex water-conducting cells defined by the presence are characterized by the presence of S-type tracheids, named of lignified cell wall thickenings (Knoll and Rothwell 1981; after the genus Sennicaulis. S-type tracheids have annular or Gensel and Andrews 1987; Knoll and Niklas 1987). helical thickenings and lateral walls that appear to be made Most attempts to understand the early evolution of tracheids of a spongy or reticulate material that may be partially deg- have centered on fossilized Silurian and Devonian vascular radation-resistant in the fossil record (fig. 2). A very thin deg- plants, and these efforts have led to a wealth of new infor- radation-resistant layer of secondary cell wall material with mation on early water-conducting cells (Grierson 1976; Zdeb- micropores appears to overlie the entire spongy layer of wall ska 1982; Kenrick and Edwards 1988; Li 1990; Kenrick and material (Kenrick and Crane 1991; Kenrick et al. 1991). Crane 1991, 1997a; Kenrick et al. 1991; Edwards et al. 1992; The eutracheophyte clade is recognized by the presence of tracheids with a relatively thicker degradation-resistant layer 1 Author for correspondence and reprints; E-mail ned@ of secondary cell wall (Kenrick and Crane 1991, 1997a). Al- colorado.edu. though much work remains to be done on the more precise Manuscript received March 1998; revised manuscript received April 1998. relationships of basal members of the eutracheophyte clade,

881 This content downloaded from 128.103.149.052 on April 12, 2016 13:08:45 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). 882 INTERNATIONAL JOURNAL OF PLANT SCIENCES

similar to G-type tracheids in possessing secondary cell wall thickenings that appear hollow (fig. 2). Although certain aspects of cell wall patterning differ among Late Silurian and Early Devonian S-, G-, and P-type tracheids, all of these early water-conducting cells possess cell wall thickenings composed of two distinct layers: a degradation-prone layer adjacent to the primary cell wall and a degradation- resistant (possibly lignified) layer next to the cell lumen (fig. 2; Kenrick and Edwards 1988; Kenrick and Crane 1991; Ken- rick et al. 1991; Edwards 1993). Developmental studies of cell wall structure in tracheary elements of extant vascular plants have been confined to highly derived seed plants (Esau et al. 1963, 1966a, 1966b; Wooding and Northcote 1964; Cron- shaw and Bouck 1965; O’Brien and Thimann 1967; Hepler et al. 1970; Esau 1978; Daniel and Nilsson 1984; Uehara and Hogetsu 1993; Fineran 1997), and it is evident that basic features of cell wall structure in tracheary elements of seed plants do not correspond well to those of S-, G-, and P-type tracheids of the Late Silurian and Early Devonian. Electron micrographs of tracheary elements in conifers and angiosperms depict cell wall thickenings with a three-layered secondary cell wall (S1, S2, and S3 layers), and these layers of secondary wall are all heavily lignified, differing mostly in the orientation (angle) of Fig. 1 Hypothesis of phylogenetic relationships among the three microfibril deposition (Boudet et al. 1995). Thus, tracheary major vascular plant lineages (Kenrick and Crane 1997a). Rhyniopsida elements of extant seed plants do not exhibit the prominent is sister to the eutracheophytes. Lycophytina, including the extant lycopsids, is the sister group to Euphyllophytina, which includes all other degradation-prone (possibly unlignified) layer of cell wall ma- extant vascular plants. Huperzia is a basal extant member of the terial that is characteristic of tracheids in early tracheophytes. Lycophytina. A connection between the hollow wall thickenings of early fossil tracheids and a possibly unlignified core in the secondary wall thickenings of basal extant vascular plants (Bierhorst two main lineages have been recognized: the Lycophytina (ly- 1958, 1960) has previously been suggested by several paleo- copsids and their extinct ancestors or close relatives, the zos- botanists (Brauer 1980; Taylor 1986; Kenrick and Edwards terophylls; Niklas and Banks 1990; Gensel 1992), whose ear- 1988; Kenrick and Crane 1991, 1997a; Kenrick et al. 1991). liest members have G-type tracheids, and the Euphyllophytina On the basis of a series of light microscope level studies of (the extinct trimerophytes, ferns, sphenopsids, psilophytes, mature tracheids, Bierhorst (1958, 1960) reported an “unlig- progymnosperms, and seed plants), whose earliest (trimero- nified or very faintly lignified” core in the interior of annular phyte) members have P-type tracheids (Kenrick and Crane or helical thickenings in Lycopodium and other basal extant vascular plants (Equisetum and Osmunda). Bierhorst (1958, 1997a). 1960) was uncertain as to whether these putatively unlignified G-type tracheids, named after the zosterophyll genus Gos- or faintly lignified wall layers consisted of primary or second- slingia, have annular or helical wall thickenings that exhibit ary walls. No photomicrographs of the reported unlignified two layers: a carbonaceous dark layer closest to the cell lumen core in the tracheid walls of primitive vascular plants were and a light layer that appears to represent a mineralized hollow published (Bierhorst 1958, 1960), so it is impossible to eval- core of each thickening (fig. 2; Kenrick and Edwards 1988; uate the specific nature of these observations. Studies of tra- Kenrick and Crane 1991; Kenrick et al. 1991). Between the cheid wall development among seedless vascular plants have spiral thickenings is a degradation-resistant layer of wall ma- not been undertaken in modern times. terial with irregularly shaped holes that range in size from less Thus, despite major progress in understanding the structural than 1 mm to ca. 4 mm (Kenrick and Edwards 1988). The diversity of early plant tracheids, interpretation of the evolu- hollow core in wall thickenings of these fossil cells may rep- tionary relationships of these fossil tracheids to those of extant resent an unlignified portion of the thickening that was pref- plants remains uncertain. S-, G- and P-type tracheids, as well erentially degraded during fossilization (Kenrick and Edwards as the tracheary elements of extant vascular plants, may all be 1988; Kenrick and Crane 1991). evolutionarily homologous and represent developmental trans- P-type tracheids are named after the genus Psilophyton (Ken- formations of a rudimentary tracheid type from a common rick and Crane 1997a), a primitive trimerophyte (Gensel ancestor of all vascular plants (Kenrick and Crane 1991, 1979). Trimerophytes are a paraphyletic assemblage of fossil 1997a). Alternatively, there may have been multiple origins of plants hypothesized to have given rise to horsetails, ferns, and water-conducting cells with secondary wall thickenings (i.e., seed plants (Banks 1975). P-type cells exhibit bordered pits “tracheids” are homoplasious; Kenrick and Crane 1991). with strands of secondary wall material that traverse the pit In order to decipher the early evolution of tracheids, devel- apertures and connect the pit borders (Gensel 1979; Hartman opmental and comparative studies of plesiomorphic basal vas- and Banks 1980). In the fossil record, P-type tracheids are cular plants are needed. Molecular phylogenetic analysis in-

Fig. 2 Longitudinal views of cell wall thickenings in fossil S-, G-, and P-type tracheids and in tracheids of extant seed plants. Early diversification of vascular plants is hypothesized to have produced an early divergent Rhyniopsida (all extinct), with S-type tracheids; the Lycophytina (lycopsids and zosterophylls), with G-type tracheids (in Late Silurian and Early Devonian members); and the Euphyllophytina (all other vascular plants), with P-type tracheids (in Early Devonian members). The cell lumen is to the right in each diagram. Cell wall thickenings in S-type cells have a thick, partially resistant spongy layer that is covered by a thin resistant layer adjacent to the cell lumen. Cell wall thickenings in G- and P-type cells appear to have a core of degradation-prone wall material that is covered by a thick resistant layer adjacent to the cell lumen. Secondary wall thickenings of extant seed plant tracheary elements appear to be homogeneous and lack any equivalent of a degradation-prone layer. dicates that extant Lycophytina are the sister group to all other Electron Microscopy extant vascular plants (Raubeson and Jansen 1992) and that For electron microscopy, specimens were postfixed2hin homosporous lycopods (Lycopodiaceae) are basal within this 2% aqueous osmium tetroxide and rinsed several times in buf- clade (Kenrick and Crane 1997a; Wikstro¨ m and Kenrick fer. Specimens for TEM were dehydrated in a graded acetone 1997). Thus, lycopods may provide critical information about series, infiltrated in three steps with Spurr’s resin, embedded, the structural features of the common ancestors of all extant and polymerized in a 70ЊC oven overnight. Thin sections (90 vascular plants. Developmental studies of tracheids in the most nm) were cut using a RMC MT-7 ultramicrotome with a Dia- basal living plants also allow us to interpret early fossil tra- tome diamond knife, stained for 15–20 min with aqueous lead cheids within a broader context and permit a deeper under- citrate, and viewed with a Phillips CM 10 transmission electron standing of the early evolution of complex water-conducting microscope at 80 kV. Specimens for SEM were dehydrated in cells in land plants. a graded ethanol series, with absolute ethanol replaced in four We examined Huperzia lucidula, a member of the most basal steps by Hemo-De clearing agent. Specimens were gradually extant clade within the Lycopodiaceae (Wagner and Beitel infiltrated and embedded in paraffin, cut into 15 mm sections 1992; Wikstro¨ m and Kenrick 1997). Huperzia is highly ple- using a rotary microtome with a disposable stainless steel siomorphic among vascular plants and can be considered a blade, and attached as ribbons to gelatin-coated coverslips. true living fossil. Morphologically, it appears very similar to After paraffin was removed, coverslips were attached to alu- the Devonian lycopods assigned to the genus Baragwanathia minum stubs with double-stick tape, sputter-coated, grounded (Hueber 1983), Drepanophycus (Li and Edwards 1995), and to the stub with silver paint, and viewed with a Zeiss DSM Asteroxylon (Hueber 1992). This is the first modern devel- 940 A scanning electron microscope. (SEM technique was opmental study of tracheid cell wall structure in a living prim- modified from Carlquist and Schneider 1997.) itive vascular plant. Enzyme Degradation Treatments Material and Methods In order to simulate possible taphonomic changes in cell wall structure associated with the fossilization process, stem Plant Material and root segments 4 mm long were immersed in an aqueous solution of 2% pectinase and 2% cellulase for 2–4 d at room Whole plants of Huperzia lucidula were obtained from Car- temperature. Segments were removed from the enzyme solu- olina Biological Supply. Stems and roots were sliced into seg- tion with fine forceps, rinsed in water, and stripped of any ments 2 mm long, fixed6hin4%glutaraldehyde in 50 mmol remaining cortex so that only the xylem cells of the vascular LϪ1 phosphate buffer (pH 7), and rinsed several times in buffer. cylinder remained. Specimens for TEM were subsequently pre- Material subjected to degradative enzyme treatment was not pared as described above. Specimens for SEM were allowed fixed. to air-dry on aluminum stubs that had been covered with dou-

This content downloaded from 128.103.149.052 on April 12, 2016 13:08:45 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). 884 INTERNATIONAL JOURNAL OF PLANT SCIENCES ble-sided sticky tape and sputter-coated, then viewed as de- secondary cell wall (fig. 5B). The later-formed layer of sec- scribed above. ondary cell wall is deposited only on the surface of the first- formed template layer and is never in direct contact with the Results primary wall. As this later-formed wall layer continues to increase in thickness and to mature, it stains more darkly (fig. Tracheid Structure and Development in Huperzia 4E). Cell autolysis occurs at the completion of tracheid differ- Differentiation of lateral walls in tracheids of Huperzia in- entiation (fig. 5C). All regions of primary cell wall not covered volves three discrete stages of cell wall deposition, each of by secondary cell wall are partially hydrolyzed (fig. 5D). Hy- which produces a cell wall layer with distinct properties. As drolysis facilitates increased water flow between cells. At ma- is true for all plants, cell wall formation is centripetal. The turity, three distinct layers of cell wall can be discerned in first layer to mature is the primary cell wall, which is smooth tracheids (both protoxylem, not shown, and metaxylem) of and homogeneous when viewed with SEM (fig. 3A). Deposi- Huperzia lucidula: a homogeneous primary cell wall; a mot- tion of the primary cell wall is completed before synthesis of tled, heterogeneous first-formed (template) layer of secondary the two layers that compose the secondary cell wall is initiated. cell wall that covers much of the primary cell wall; and a In Huperzia, a first-formed layer of secondary cell wall homogeneous layer of secondary cell wall that overlies the (dark, mottled appearance under the TEM) is deposited over template layer (fig. 4F). the surface of the primary cell wall, except in areas through The distinct nature of the two layers of secondary cell wall which water will flow (pits; figs. 3B;4A, B). This wall layer in Huperzia extends beyond the ultrastructural level. When is deposited by distinctive Golgi-derived vesicles that contain prepared for SEM, secondary thickenings sometimes pulled electron-opaque substances surrounded by electron-lucent ma- apart (fig. 6A), presumably breaking at the interface between terial (fig. 5A). The first-formed secondary cell wall layer ap- the first-formed and the later-formed wall layers. In some cases, pears to determine the pattern of further secondary cell wall the first-formed layer of secondary wall remained with the deposition. For this reason, we call it the “template layer.” primary wall when the rest of the thickening had been pulled After completion of the template layer, an additional and away (fig. 6B). structurally distinct layer of secondary cell wall (“resistant layer”—see below) is deposited. This later-formed layer is first Response of Huperzia Tracheids to discernible next to the cell lumen as a thin, very lightly stained Degradative Treatment layer of newly synthesized cell wall material (fig. 4C, D). Golgi- derived vesicles that deposit the later-formed layer of secondary The two layers of secondary cell wall in tracheids of Hu- cell wall material do not contain the electron-opaque sub- perzia are distinct chemically as well as structurally and de- stances associated with deposition of the template layer of the velopmentally. Stem segments of Huperzia were subjected to

Fig. 3 Longitudinal views of mature (A) and developing (B) metaxylem tracheids in Huperzia. A, In cell on right, secondary cell wall (SW) has been pulled away, exposing the smooth and homogeneous primary cell wall (PW). Bar ϭ 5 mm. B, Cell in center is in face view, demonstrating that the ﬁrst-formed layer of secondary wall (“template layer,” T) covers the entire primary wall except for pits (P). Bar ϭ 1 mm.

Fig. 4 Development of secondary cell wall thickenings in metaxylem tracheids of Huperzia (longitudinal views). A and B, First-formed layer of secondary wall (“template layer”; T) is deposited on the primary cell wall (P). C–E, Subsequently a later-formed layer of secondary wall (“resistant layer”; R) is deposited on the surface of the template layer. C, D, The resistant layer appears unstained or very lightly stained when ﬁrst deposited. E, Subsequently, the resistant layer appears gray. F, In mature (dead) tracheids, the template layer remains distinct from the resistant layer of secondary cell wall. A dark layer of recently lysed cytoplasm (Cy) coats the secondary wall. Bar (B) for whole figure ϭ 0.3 mm.

enzyme treatment in order to simulate possible taphonomic degrading enzymes than is the later-formed layer of secondary changes in cell wall structure that may have been experienced wall located next to the cell lumen. Because the first-formed by tracheids found in the fossil record. After being immersed secondary cell wall layer in Huperzia appears to determine the in a mixture of the cell wall degrading enzymes pectinase and pattern of further secondary wall deposition, we have called cellulase, the stem cortex was severely degraded, so that easy it the “template layer.” Because the later-formed secondary isolation of the vascular cylinders was possible. The vascular wall layer is resistant to wall degrading enzymes, we have cylinder remained intact because of an enzyme-resistant en- called it the “resistant layer.” dodermoid sheath surrounding it, but the only cells that re- Almost all studies of tracheary element differentiation and mained inside the sheath were tracheids with resistant wall fine structure have been conducted on seed plants, namely, compounds. conifers or highly derived flowering plants (Esau et al. 1963, In tracheids from control stems that were not exposed to 1966a, 1966b; Wooding and Northcote 1964; Cronshaw and enzymatic treatment, the primary wall and both the template Bouck 1965; O’Brien and Thimann 1967; Hepler et al. 1970; layer and the later-formed layer of secondary wall are present Esau 1978; Fineran 1997). In angiosperms that have been stud- and intact (fig. 7A, C). Tracheids from stem segments that were ied, lignification begins at the cell periphery and gradually immersed in a solution of pectinase and cellulase showed con- progresses toward the cell lumen as centripetal wall deposition siderable alterations in the appearance of the cell walls. The progresses (Hepler et al. 1970; Boudet et al. 1995). Electron primary cell wall and the first-formed layer of secondary cell micrographs of mature tracheary elements in conifers and an- wall in treated tracheids were either entirely degraded (missing) giosperms depict secondary cell wall thickenings that are ho- or were represented by a delicate reticulum of remaining cell mogeneous and lignified throughout (Esau 1977; Boudet et al. wall material (fig. 7B, D). Only the later-formed secondary 1995). Thus, the secondary cell walls of tracheary elements of cell wall layer (next to the cell lumen) was not degraded by conifers and angiosperms do not display a distinctive two- enzyme treatment. For this reason, we refer to this portion of layered structure that corresponds to wall thickenings of fossil the secondary wall as the “resistant layer.” S-, G-, and P-type tracheids and the tracheids of extant Huperzia. Discussion Our cell wall degradative studies suggest that spatial distribution of lignin in Huperzia may be heterogeneous. In Hu- perzia, lignin may be predominantly localized in the resistant layer of the secondary cell wall adjacent to the cell lumen and This is the first modern developmental study of tracheid largely or entirely absent from the degradation-prone template secondary wall structure in a primitive vascular plant. In the layer. However, it is possible that different types of lignin chem- lycopod Huperzia lucidula, a distinctive first-formed layer of istry (monomer content, degree of polymerization, or inter- secondary wall located next to the primary wall at the cell actions between lignin and other cell wall compounds) may periphery is described. A later-formed secondary wall layer is distinguish the template and resistant layers of tracheids in deposited over the first-formed layer. Tracheids of Huperzia Huperzia. were subjected to degradative treatments to simulate condi- Because different layers of cell walls in tracheids are difficult tions experienced by cells in the fossil record. The first-formed to distinguish at the light microscope level, determination of layer of secondary cell wall is far less resistant to cell wall the spatial distribution of lignin within a developmental con-

Fig. 5 Development and structure of Huperzia tracheids. A, Golgi vesicles (arrows) depositing the template layer (T) contain electron-opaque substances. Bar ϭ 0.2 mm;BV ϭ budding vesicle;FV ϭ fusing vesicle. B, Electron-opaque substances are not found in Golgi vesicles (arrows) associated with deposition of the resistant layer (R). Bar ϭ 0.4mm. C, D, Cross-sectional views of developing and mature tracheids. Bars ϭ 1 mm. C, Developing tracheids. Cell in center with dark cytoplasm is beginning to undergo autolysis. D, Mature, dead tracheids. Hydrolyzed primary walls (arrows) facilitate lateral transport of water between cells.

text will require transmission electron microscope level reso- another method for understanding the relationship between lution. Lignins are complicated heteropolymers, and the in- ligniﬁcation, wall structure, and degradation properties. Re- teractions between lignin monomers and other cell wall cently, reports of successful immunolocalization of lignin in molecules are complex (Whetten and Sederoff 1995). Never- angiosperms (Zea and Triticum) have been published (Ruel et theless, cytochemical tests for lignin in situ within cell walls al. 1994; Burlat et al. 1997; Joseleau and Ruel 1997; Mu¨ sel at the ultrastructural level have sometimes been successful (Fi- et al. 1997). When assays for in situ lignin content become neran 1997). An immunological approach to localization of more widely available, it may be possible to more precisely lignin in speciﬁc regions of the cell walls of tracheids represents analyze the cell wall chemistry of the template and resistant

layers of secondary cell wall thickenings in the tracheids of Huperzia.

Comparisons of Huperzia Tracheids with Early Fossil Tracheids An equivalent of the template layer in Huperzia has not been described in tracheary elements of any other extant vascular plants. The position of the template layer in Huperzia, and its susceptibility to degradation, strongly suggest developmental and structural homology with the degradation-prone wall layer in the cell wall thickenings of tracheids characteristic of early vascular plants. As a number of paleobotanists have noted, the absent layer of wall in G-type fossil tracheids may represent an unligniﬁed or poorly ligniﬁed wall layer that has not survived the fossilization process (Brauer 1980; Taylor 1986; Kenrick and Ed- wards 1988; Kenrick et al. 1991; Kenrick and Crane 1991, 1997a). The degradation-prone template layer of Huperzia corresponds precisely to the missing core of cell wall thick- Fig. 6 Sectional view (A) and face view (B) of secondary wall thickenings in G-type fossil tracheids. The resistant layer of Hu- enings that have pulled apart during preparation for SEM, presumably perzia is positionally matched with the degradation-resistant at the interface between the template layer (T) and the resistant layer cell wall layer of G-type cells. Structural correspondence of (R). The template layer remains with the primary cell wall. Bars ϭ secondary cell wall layers of Huperzia tracheids extends be- 1 mm. yond the ancestral G-type cells of the Lycophytina to the P- type cells representative of the earliest members of the Eu- phyllophytina (the sister group to Lycophytina). Cell wall

Fig. 7 Comparison of wall layers in tracheids of Huperzia treated with wall-degrading enzymes and control untreated tracheids. A, C,In control (untreated) cells, the primary wall (PW), the ﬁrst-formed (template) layer of secondary wall (T), and the later-formed (resistant) layer of secondary wall (R) are intact. B, D, In cells subjected to enzyme treatment, the template layer (T) has been degraded and appears reticulate, while the resistant layer (R) remains intact. Bars: A, B ϭ 0.3 mm; C, D ϭ 1.5 mm.

Fig. 8 Hypothesis for evolution of tracheid development and lateral wall structure. A, B, Secondary walls are added to smooth primary walls, creating cells with a partially resistant, differentially thickened template layer of secondary wall. C, Deposition of a thin resistant layer of cell wall on the surface of the template layer produces a wall thickening at an early stage of tracheid development in Huperzia that is similar to those of mature S-type fossil tracheids. D, Continued deposition of resistant wall material produces a secondary thickening at an intermediate stage of development in tracheids of extant Huperzia that is similar to mature G- and P-type fossil tracheids. E, Mature tracheids of extant Huperzia have secondary cell wall thickenings that are approximately half template layer and half resistant layer. F, Continued reduction of the template layer results in the characteristic secondary cell wall thickenings of seed plants, in which the entire secondary wall is composed of resistant material.

thickenings of fossil P-type cells may also have an absent core fossil vascular plants, developmental data from Huperzia dem- layer that positionally matches the template layer of Huperzia onstrate a clear structural correspondence between the tra- and underlies a resistant layer (Hartman and Banks 1980; Ed- cheids of extinct and extant primitive vascular plants. More- wards 1993; Kenrick and Crane 1997a, 1997b). over, tracheid development in Huperzia provides the essential Our developmental analysis and enzyme degradation ex- information needed to propose an explicit model for the ev- periments also suggest a connection between the tracheids of olution of secondary wall thickenings in tracheids of vascular Huperzia and the more primitive fossil S-type cells. In G- and plants (fig. 8). P-type tracheids, the layer that we propose is homologous to the template layer of Huperzia is frequently absent in the fossil Developmental Model for the Early Evolution of Tracheids record, whereas this layer in S-type cells is somewhat resistant to degradation. The spongy portion of S-type cell wall thick- We hypothesize that the evolution of secondary cell wall enings may have survived in the fossil record because it was thickenings in tracheids began with deposition of a partially partially resistant to degradation, perhaps as a result of thin degradation-resistant template layer of cell wall material onto veins of degradation-resistant biopolymers running through a the lumen surface of the primary cell wall (fig. 8A, B). Tra- mass of less resistant material (Kenrick and Crane 1991; Ken- cheids bearing a degradation-prone template layer (as in S- rick et al. 1991). Similarly, the template layer of Huperzia does type tracheids), but lacking a resistant layer, have not been not completely disappear under conditions of degradation but described in the fragmentary fossil record; but we predict that instead assumes a reticulate condition that resembles the eventually such cells will be found (e.g., the poorly preserved spongy wall layer of S-type cells (fig. 7B, D). Thus, the template specimens of Horneophyton and Langiophyton described by layer of Huperzia may be homologous to the spongy layer of Kidston and Lang [1920] and Remy et al. [1993] appear to the primitive S-type cells. The apparent differences between show weakly developed thickenings in water-conducting cells). the degradation-prone layers of fossil G- and P-type tracheids A further innovation, a thin resistant layer, was added sub- and those of fossil S-type cells (and Huperzia) suggest that sequently to the process of secondary cell wall deposition (fig. chemical diversification of the template layer may have oc- 8C). This condition is found in mature S-type conducting cells curred during the early evolution of vascular plants. In addi- of the most primitive vascular plants, the Rhyniopsida. Later tion, the resistant layer of Huperzia may be homologous not modification of the process of tracheid wall deposition led to only to the thick resistant layer of G- and P-type cells but also a decrease in the thickness of the template layer and an increase to the thin resistant cell wall layer next to the cell lumen in S- in the thickness of the resistant layer, as found in G- and P- type cells. type fossil tracheids (fig. 8D). Our documentation of a degradation-prone template layer A thick resistant tracheid wall layer may be a synapomorphy and a resistant layer in tracheid secondary cell wall thickenings of the eutracheophytes, that is, Lycophytina plus Euphyllo- of the primitive extant plant Huperzia provides the critical link phytina (fig. 1; Kenrick and Crane 1997a) or may have evolved between the water-conducting cells of Late Silurian and Early independently in these two lineages. Several researchers have Devonian tracheophytes and those of their extant descendants. noted a trend toward increasing thickness of the resistant layer While tracheid cell wall thickenings of previously studied ex- in early fossil tracheids (Kenrick and Edwards 1988; Kenrick tant plants appear to differ significantly from those of early and Crane 1991, 1997a). We have calculated that a resistant

This content downloaded from 128.103.149.052 on April 12, 2016 13:08:45 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c). COOK & FRIEDMAN—ORIGIN AND EARLY EVOLUTION OF TRACHEIDS 889 layer composes ca. 2% of the thickness of secondary wall vascular plant, Huperzia. Precise correspondence between the thickenings in S-type cells, 30% in G- and P-type cells, and patterning of tracheid wall layers in Huperzia and the earliest 50% in extant Huperzia. We propose that a continuation of fossil vascular plants also provides support for the hypothesis the trend toward reduction of the template layer and aug- of a single origin of tracheids in land plants. mentation of the resistant layer has produced the secondary cell wall thickenings characteristic of tracheids of extant seed Acknowledgments plants (ﬁg. 8D, E), in which no equivalent of a template layer that contains electron-opaque particles has been reported (see We thank P. Diggle, J. Hanken, P. Kenrick, and A. Knoll for Esau et al. 1966a) and the entire secondary cell wall is highly critical reading of the manuscript and P. Gensel for stimulating degradation-resistant. discussions about this research. This work was supported in In summary, we have discovered a previously unknown pat- part by the National Science Foundation (IBN-9696013) and tern of secondary cell wall formation in tracheids of extant grants-in-aid-of-research from Apple Computer, Lasergraph- vascular plants. We demonstrate that the earliest fossil tra- ics, Carl Zeiss, Research and Manufacturing Company, Leica cheids are now interpretable within the context of develop- Instruments, Fisher Scientiﬁc, and Olympus America to W. E. mental information derived from an extremely primitive extant Friedman.

Literature Cited

Banks HP 1975 Reclassification of Psilophyta. Taxon 24:410–413. copsids: evidence from morphology, paleoecology, and cladistic Bierhorst DW 1958 The tracheary elements of Equisetum with ob- methods of inference. Ann Mo Bot Gard 79:450–473. servations on the ontogeny of the internodal xylem. Bull Torrey Bot Gensel PG, HN Andrews 1987 The evolution of early land plants. Club 85:416–433. Am Sci 75:478–489. ——— 1960 Observations on tracheary elements. Phytomorphology Gray J 1993 Major Paleozoic land plant evolutionary bio-events. Pa- 10:249–305. laeogeogr Palaeoclimatol Palaeoecol 104:153–169. Boudet AM, C Lapierre, J Grima-Pettenati 1995 Biochemistry and Gray J, D Massa, AJ Boucot 1982 Caradocian land plant microfossils molecular biology of lignification. New Phytol 129:203–236. from Libya. Geology 10:197–201. Brauer DF 1980 Barinophyton citrulliforme (Barinophytales incertae Grierson JD 1976 Leclercqia complexa (Lycopsida, Middle Devo- sedis, Barinophytaceae) from the Upper Devonian of Pennsylvania. nian): its anatomy, and the interpretation of pyrite petrifications. Am J Bot 67:1186–1206. Am J Bot 63:1184–1202. Burlat V, K Ambert, K Ruel, J-P Joseleau 1997 Relationship between Hartman CM, HP Banks 1980 Pitting in Psilophyton dawsonii,an the nature of lignin and the morphology of degradation performed Early Devonian trimerophyte. Am J Bot 67:400–412. by white-rot fungi. Plant Physiol Biochem 35:645–654. Hepler PK, DE Fosket, EH Newcomb 1970 Lignification during sec- Carlquist S, EL Schneider 1997 SEM studies on vessels in ferns. 2. ondary wall formation in Coleus: an electron microscopic study. Am Pteridium. Am J Bot 84:581–587. J Bot 57:85–96. Cronshaw J, G Bouck 1965 The fine structure of differentiating xylem Hueber FM 1983 A new species of Baragwanathia from the Sextant elements. J Cell Biol 24:415–431. Formation (Emsian), northern Ontario, Canada. Bot J Linn Soc 86: Daniel G, T Nilsson 1984 Studies on the S2 layer of Pinus sylvestris. 57–79. Report no. 154. Swedish University of Agricultural Sciences, De- ——— 1992 Thoughts on the early lycopsids and zosterophylls. Ann partment of Forest Products, Uppsala. Mo Bot Gard 79:474–499. Edwards D 1993 Cells and tissues in the vegetative sporophytes of Joseleau J-P, K Ruel 1997 Study of lignification by noninvasive tech- early land plants. New Phytol 125:225–247. niques in growing maize internodes. Plant Physiol 114:1123–1133. Edwards D, KL Davies, L Axe 1992 A vascular conducting strand in Kenrick P, PR Crane 1991 Water-conducting cells in early fossil land the early land plant Cooksonia. Nature 357:683–685. plants: implications for the early evolution of tracheophytes. Bot Esau K 1977 Anatomy of seed plants. Wiley, New York. Gaz 152:335–356. ——— 1978 On vessel member differentiation in the bean (Phaseolus ——— 1997a The origin and early diversification of land plants: a vulgaris L.). Ann Bot 42:665–677. cladistic study. Smithsonian Institution Press, Washington. Esau K, VI Cheadle, RH Gill 1966a Cytology of differentiating tra- ——— 1997b The origin and early evolution of plants on land. Na- cheary elements. I. Organelles and membrane systems. Am J Bot ture 389:33–39. 53:756–764. Kenrick P, D Edwards 1988 The anatomy of Lower Devonian Gos- ——— 1966b Cytology of differentiating tracheary elements. II. slingia breconensis heard based on pyritized axes, with some com- Structures associated with cell surfaces. Am J Bot 53:765–771. ments on the permineralization process. Bot J Linn Soc 97:95–123. Esau K, VI Cheadle, EB Risley 1963 A view of ultrastructure of Cu- Kenrick P, D Edwards, RC Dales 1991 Novel ultrastructure in water- curbita xylem. Bot Gaz 124:311–316. conducting cells of the Lower Devonian plant Sennicaulis hippo- Fineran BA 1997 Cyto- and histochemical demonstration of lignins crepiformis. Palaeontology 34:751–766. in plant cell walls: an evaluation of the chlorine water/ethanolamine- Kidston R, WH Lang 1920 On old red sandstone plants showing silver nitrate method of Coppick and Fowler. Protoplasma 198: structure, from the Rhynie chert bed, Aberdeenshire. II. Additional 186–201. notes on Rhynia gwynne-vaughani, Kidston and Lang: with descrip- Gensel PG 1979 Two Psilophyton species from the Lower Devonian tions of Rhynia major, n. sp., and Hornea lignieri, n. g., n. sp. Trans of eastern Canada with a discussion on morphological variation R Soc Edinb 52:603–627. within the genus. Palaeontographica B168:81–99. Knoll AH, KJ Niklas 1987 Adaptation, plant evolution, and the fossil ——— 1992 Phylogenetic relationships of the zosterophylls and ly- record. Rev Palaeobot Palynol 50:127–149.

Knoll AH, GW Rothwell 1981 Paleobotany: perspectives in 1980. Ruel K, O Faix, J-P Joseleau 1994 New immunogold probes for stud- Paleobiology 7:7–35. ying the distribution of the different lignin types during plant cell Li C-S 1990 Minarodendron cathaysiense (gen. et comb. nov.), a ly- wall biogenesis. J Trace Microprobe Tech 12:247–265. copod from the late Middle Devonion of Yunnan, China. Palaeon- Taylor DW 1986 Anatomical and morphological study of a new spe- tographica B220:97–117. cies of Taeniocrada: a Devonian tracheophyte from New York State. Li C-S, D Edwards 1995 A re-investigation of Halle’s Drepanophycus Rev Palaeobot Palynol 47:63–87. spinaeformis Gopp. from the Lower Devonian of Yunnan Province, Uehara K, T Hogetsu 1993 Arrangement of cortical microtubules southern China. Bot J Linn Soc 118:163–192. during formation of bordered pits in the tracheids of Taxus. Pro- Mu¨ sel G, T Schindler, R Bergfeld, K Ruel, G Jacquet, C Lapierre, V toplasma 172:145–153. Speth, P Schopfer 1997 Structure and distribution of lignin in pri- Wagner WH Jr, JM Beitel 1992 Generic classiﬁcation of modern mary and secondary cell walls of maize coleoptiles analyzed by North American Lycopodiaceae. Ann Mo Bot Gard 79:676–686. chemical and immunological probes. Planta 201:146–159. Whetton R, R Sederoff 1995 Lignin biosynthesis. Plant Cell 7: Niklas KJ, HP Banks 1990 A reevaluation of the Zosterophyllophy- 1001–1013. tina with comments on the origin of lycopods. Am J Bot 77:274–283. Wikstro¨ m N, P Kenrick 1997 Phylogeny of Lycopodiaceae (Lycop- O’Brien TP, KV Thimann 1967 Observations on the ﬁne structure of the oat coleoptile. III. Correlated light and electron microscopy of sida) and the relationships of Phylloglossum drummondii Kunze the vascular tissues. Protoplasma 63:443–478. based on rbcL sequences. Int J Plant Sci 158:862–871. Raubeson LA, RK Jansen 1992 Chloroplast DNA evidence on the Wooding FBP, DH Northcote 1964 The development of the second- ancient evolutionary split in vascular land plants. Science 255: ary wall of the xylem in Acer pseudoplatanus. J Cell Biol 23: 1697–1699. 327–337. Remy W, PG Gensel, H Hass 1993 The gametophyte generation of Zdebska D 1982 A new zosterophyll from the Lower Devonian of some early Devonian land plants. Int J Plant Sci 154:35–58. Poland. Palaeontology 25:247–263.

This content downloaded from 128.103.149.052 on April 12, 2016 13:08:45 PM All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).