Phylogeny Of Subfamily

Verbenoideae (Verbenaceae)

 Based On

nrDNA Sequences.


Jorge Ruiz


Dr. Ed C. Freeman

Phylogeny of subfamily Verbenoideae (Verbenaceae) based on nrDNA sequences.


The family Verbenaceae is essentially a pantropical taxon with a few genera which reach the warm temperate, and even the cool temperate regions (Lawrence, 1951).  It is composed of about 75 genera and 3000 species (Smith 1977). This makes the Verbenaceae a rather large family. They usually are herbs, shrubs, or trees.  The flowers are bisexual, typically zygomorphic, small to medium-sized, usually very irregular or regular to somewhat irregular and 5-merous (occasionally 4-merous or 8-merous, and their stems are often four-sided (Lawrence, 1951; Smith 1977). Other characteristics of the Verbenaceae are that they are either anatopous ovules (micropyle close to funiculus attachment) or orthotropous ovules (erect and straight) (Smith 1977).

The Verbena Family is has commonly been accepted as belonging to an old Order (a name not commonly recognized today), Tubiflorae, with a close affinity to the Lamiaceae (or Labiatae) (Lawrence, 1951).  Currently, most authorities consider Verbenaceae to be within the order Lamiales along with the families Borgarinaceae (the borage family with 100 genera and 2,000 species), Callitrichaceae (the water starwort family with 1 genus and 25 species), Phrymaceae (the phryma family with 1 genus and 2 species), and Lamiaceae (the mint family with 180 genera and 3,500 species) (Smith 1977).  There has been no consensus agreement as to the circumscription of the Verbenaceae.  Bessey (1948) separate the Lamiaceae and Verbenaceae (on the basis of corolla zygomorphy and gynoecial characters) as a distinct order, Lamiales (Lawrence, 1951).  Hallier retained it with in the Tubiflorae and derived it from the figwort family (Scrophulariaceae) (Lawrence, 1951).  Hutchinson broadened its circumscription to include the Phyrmaceae and (as of 1926) including it with in the Lamiales (Lawrence, 1951).   Later, in 1948, he placed them in the Verbenales (Lawrence, 1951).  Hutchinson also considered the Verbenaceae to be unrelated to the Lamiaceae and posited that they arose from the Rubiaceae (Lawrence, 1951).  Further, some genera of the classical family Verbenaceae have been placed in the Lamiaceae.  It is clear, then, that not only is the evolution of Verbenaceae in question but even its delimitation and origins and closest relatives.

The family has been treated several ways in the literature.  Some authorities (e.g. Smith 1977) consider it to be composed of three subfamilies.  The first is the Verbenoideae with racemose or spicate inflorescences.  This subfamily is made up of the genera Verbena, Glandularia, Lantana, Phyla, Stylodon, Priva, and Duranta.  The second subfamily in this taxonomic scheme is the Vitioicoideae with cymose inflorescences. This subfamily is composed of the genera Callicarpa, Tectona, Clerodendrum, and Vitex.  The third subfamily, Avicennoideae, is composed of the genus Avicennia (the black mangroves of coastal Louisiana to Florida in the United States). Many authorities have segregated to genus Avicennia into the family Avicenniaceae. Circumscribing the family has been problematic. Some genera (e.g., Callicarpa, Clerodendrum, and Vitex) have been considered by some to be members of the Lamiaceae (the mint family).

In this project we are trying to determine the evolutionary relationships of genera within the Verbenaceae (subfamily Verbenoideae) based on nucleic acid sequence similarities in nuclear ribosomal DNA sequences.  In this study we will be using non-coding DNA sequences which presumably accumulate nucleic acid base changes faster than coding sequences. We will determine the sequences of the ITS 1 and ITS 2 regions (Internal Transcribed Spacers between the 18s, 5.8s, and 26s genes).  The most recent similarities are assumed to be shared derived characters that are not present in their distant ancestors and are assumed to indicate common ancestry.   Thus, the pattern of nucleic acid sequence similarities will be used to imply the genetic relationships within and between species and genera.  We will use the classic methodologies of cladistics to do this.  These will include at least maximum parsimony, neighbor-joining, quantet-puzzling algorithms.   The hypothesized relationships (which are shown as branched trees or cladograms) will be based on symplesioporphies (primitive characters which are shared by two or more taxa), synapomorphies (a derived characters that are shared by two or more taxa, and held to reflect common ancestry, and autopomorphies (a derived character unique to that single taxon).  In these methodologies, the algorithms build a progressive bifurcation of a linage where two taxa are more closely related to each other if they share a more recent common ancestor. We create a series of hypothesized relationships that can be represented in a branched diagram called a cladogram.  The algorithms usually construct numerous trees but the shortest or simplest are considered to be the most probable.

The goal is, therefore, to create a cladogram that will show taxa that share common ancestors, and their descendants (monophyletic groups). We wish to avoid polyphyletic assemblages (in which taxa are placed together even though they do not share a common genetic lineage) in a classification scheme.


We will be using different protocols of DNA extraction and purification from leaves since no single extraction method is equally effective on all leaf material.  For example, some leaves contain quantities of secondary metabolites like polyphenolics and/or polysaccharides which can interfere with the PCR chemistry.  These compounds coprecipitate with DNA and are difficult to remove from an extract.  Consequently, an extraction protocol usually has to be optimized for each species. 

A specimen of dry or fresh leaf material (approximately 10 mg of dry leaf or 50 mg of fresh leaf tissue is ground into a fine powder.  Fresh leaf material can be ground in either liquid nitrogen or dry material can be ground in a mortar and pestle.  To the leaf power is added about 500 µL of a lysis buffer like CTAB solution (e.g. Doyle and Doyle 1987) or extracted using a proprietary lysis buffer such as DNAzol™ (Molecular Research Center, Cincinnati OH) with its protocol or the MasterPure™ plant DNA extraction kit (Epicenter Inc., Madison WI) in a 1.5 ml microfuge tube. The tube contain the tissue is placed in a heating block at 65ºC to accelerate the destruction of the cell organelle membranes by detergents.  To the extracts from dry leaf samples we will add proteinase K to digest proteins during the membrane degredation step.  After incubating for at least 30 min. at 65C, the pigments and remaining proteins were removed using chloroform/isoamyl-alcohol (24:1 v/v).  This solution is called CIAA solution.  The aqueous solution is then mix well with an equal volume of CIAA and centrifuged for 5-10 min.  at 14,000 X g. The photosynthetic pigments go into solution in the CIAA (the lower phase in the tube after centrifugation). The cell debris and proteins form a layer at the aqueous solution/CIAA interface.  The aqueous phase above contains the DNA.   The aqueous layer is transferred to a new 1.5 mL microfuge tube. Isopropanol (an equal volume) is then added and thoroughly mixed to precipitate the DNA.  The tube is then centrifuged for about 2-5 minutes. A white to clear pellet can then seen at the bottom of the tube.   The pellet is then washed with 70% ethanol to remove water-soluble contaminates without allowing the DNA to go into solution.  A brief centrifugation follows and the 70% ethanol solution is decanted.   Another brief centrifugation follows to remove remaining 70% ethanol with a pipetter.    The pellet is then dried.  The dried pellet is re-dissolved in water or TE buffer.  At this time RNase can be added and incubated for about 15 min. at room temperature to eliminate RNA in the sample which co-precipitated with DNA.  An aliquot of the extracted DNA is then electophoresed in a 0.8% HMP agarose gel to determine whether DNA is present and the degree of fragmentation.  To the gel, a small amount ethidium bromide is added.  Ethidium bromide intercalates into the double helix of DNA molecules and fluoresces when exposed to UV light. This allows the visualization of DNA in the gel.  If DNA is present and appears to be of sufficient quality, a PCR reaction is attempted to amplify the ITS region.  Primers 5* (J. Janovec, personal comm.)  and 4 of White et al. 1990) were commonly used.  Again, an aliquot of the PCR reaction is electrophoresed in a 0.8% HMP agarose to determine whether or not amplification occurred and to assess the quantity of product. 

If the PCR product is of sufficient quality and quantity, the product is then directly sequenced.  We will be using the Sequitherm XL™ II sequencing kit (Epicentre Technologies, Madison, WI) with its protocol.  The reagents are to be mixed with the template DNA and fluorescent-labeled sequencing primers (Licor Corp., Lincoln, NE).  Then, a Mastercycler thermocycler will be used to create the actual sequences.  The profile of the sequencing reaction is 10 cycles of a three-step program.  This is: (1) one minute at 98C, which is the denaturing step, (2) the temperature is then reduced to 40C at a rate of 0.5C/sec., which is the primer annealing step, and (3) the temperature is increased to 65C for five min., which is the extension step.  An aliquot of 3.0 µL of stop solution is then added to each tube.  The sequencing reactions, after denaturation at 95C for 3 min., are then loaded onto a 66 cm 3.5% polyacrylimide gel and sequenced using a Licor 4200 sequencer and electrophoresed for approximately 4 h.  The sequences are to be manually read.   The sequences are recorded, aligned, and a Nexus file containing the aligned sequences of all taxa created.  PAUP 4.0 algorithms will create the cladograms using the algorithms and conditions specified.  The algorithms to be used for cladogram generation are still undecided since the methodology is typically determined a postori.      


            So far we have been able to accomplish successful DNA extractions, PCR amplifications, and sequencing reactions on about 16 taxa (some are incomplete and need to be re-sequenced for conformation).  We still need sequence data from about 10 taxa.  However, these taxa have proved to be recalcitrant and classic DNA extraction protocols have failed to yield DNA of sufficient purity to allow PCR amplification. Extraction protocol optimization usually is the principle problem faced in sequencing plant DNA and this problem is currently delaying completion of this study. 

PCR products need to be sequenced about three times for confidence in the accuracy of the sequences.  Thus far, Aloysia gratissima, Aloysia macrostachya, Aloysia wrightii have met this test.  We are current sequencing and resequencing Citharexylem spinosum, Glandularia bipennitifida Glandularia wrightii, Phyla nodiflora, Stylodon carneus, Verbena halei, Verbena bonariense, Verbena bracteata, and Verbena rigida.  Lantana camara, Callicarpa dichotoma, and Clerodendrum speciosisimma ITS sequences have been taken from GenBank.  An ITS sequence of Lantana horrida was obtained from Raul Gutierrez, a graduate student in this lab.



Doyle, J. J., and J. L. Doyle.  1987.  A rapid DNA isolation procedure for small quantities

of fresh leaf tissue.  Phytochemical Bulletin 19:11-15.

Lawrence, G. H. M.  151.  Taxonomy of vascular plants.  McMillan Co., New York.

Smith, J. P., Jr.  1977.  Vascular plant families.  Mad River Press, Inc.: Eureka.

White, T. J., T. Bruns, S. Lee, and J. Taylor.  1990. Amplification and direct sequencing

            of fungal ribosomal RNA genes for phylogenetics. In  L. Phillips and I. K. Vasil

            (eds.), PCR Protocols.  Academic Press: New York.


State Reports | UT System | Customer Service Statement | Site Feedback | Required Links
500 West University Avenue | El Paso, Texas 79968 | Dean's Office: (915) 747-5536 | Advising Office: (915) 747-8027