Hyptidinae, ca. 400 species, is an important component of Neotropical vegetation formations. Members of the subtribe possess flowers arranged in variously modified bracteolate cymes and nutlets with an expanded areole and all share a unique explosive mechanism of pollen release, except for Asterohyptis. In a recent phylogenetic study, the group had its generic delimitations rearranged with the recognition of 19 genera in the subtribe. Although the previous phylogenetic analysis covered almost all the higher taxa in the subtribe, it lacked a broader sampling at the species level. Here we present a new expanded phylogenetic analysis for the subtribe comprising 153 accessions of Hyptidinae sequenced for the nuclear nrITS, nrETS, and waxy regions and the plastid markers trnL-F, trnS-G, trnD-T, and matK. Our results widely support the previous phylogenetic results with some changes in the support and relationship between genera. It also uncovers the need for a new combination of Eriope machrisae in Hypenia and the phylogenetic position of Hyptis sect. Rhytidea, which was demonstrated to be part of Mesosphaerum. The generic delimitation in Hyptidinae is discussed, and we recommend that further studies with more markers are needed to confirm the monophyly of Hyptidendron and Mesosphaerum, as well as to support taxonomic changes on the infrageneric delimitation within Hyptis s. s.

DNA Amplification and Sequencing—Total genomic DNA was extracted from fresh or silica-gel dried leaf material and sodium chloride/CTAB preserved material (Chase and Hills 1991) or from herbarium specimens. The fragments, when extracted from herbarium, were removed from HUEFS and K collections. A rescaled version of the Doyle and Doyle (1987) protocol was used for genomic DNA extractions. We chose the nuclear ribosomal internal transcribed spacer region (nrITS) including both ITS1 and ITS2 and intervening 5.8S and nuclear ribosomal external transcribed spacer region nrETS, and the nuclear low copy waxy granule-bound starch synthase I (GBSS). The plastid regions used were 3’trnK-matK (including partial trnK intron and matK coding region), trnL-F region (including the trnL intron and the trnL-trnF intergenic spacer), trnD-T and trnS-G region (including trnS-psbZ intergenic spacer, partial sequence; psbZ gene, complete cds; psbZ-trnG intergenic spacer, complete sequence; and tRNA-Gly (trnG) gene). The nrITS region was amplified using the primers 17SE and 26SE of Sun et al. (1994). The 3′ 18S-IGS primer of Baldwin and Markos (1998) and the 5′ primer ETS-B (Beardsley and Olmstead, 2002) were used to amplify a portion of the 3′ end of the nrETS. The nuclear region waxy were sequenced between GBSSI bd9f-bd11r, using the same methodology designed by Drew and Sytsma (2013). Therefore, for the GBSSI gene, we used a nested PCR approach to amplify the region between (and including parts of) exons 7–11. The initial PCR reaction was used the primers bd7f and bd12r Drew and Sytsma (2013). The PCR product from the above amplification was then used (after 1: 20 dilution) as a template for the additional PCR reaction, using the primers bd9f and bd11r. The product of this amplification was then sequenced with the same primers used in the nested PCR. The partial matK/trnK locus was amplified using 390F and 1326R (Cuénoud et al. 2002). The whole trnL-trnF region was amplified using primers “c” and “f”, with the use of internal primers “d” and “e” for some problematic samples, of Taberlet et al. (1991). For amplifying the trnS-G spacer we used the set of primers described in Shaw et al. (2007). The spacer trnD-T was amplified for most taxa using the primers of Demesure et al. (1995), trnD GUC and trnT GGU. For some samples which could not be amplified using these primers, we used the internal primer trnY GUA (Shaw et al. 2005). The plastid loci were sequenced using the same set of primers used for the amplification, whereas the nuclear nrITS was sequenced using internal primers ITS92 (Desfeux and Lejeune 1996) and ITS4 (White et al. 1990) with the same PCR program.

All PCR amplifications were performed in a final volume of 10 µL containing: 5 µL of TopTap master mix kit (Qiagen, Valencia, California), 2.25 pMol primers each, 5–10 ηg of genomic DNA, and ultrapure H₂O (enough to complete the volume to 10 µL). For the ITS amplification, we added 2% DMSO (dimethyl sulfoxide) and 1M of betaine. All regions were amplified using initial denaturation at 94°C (5 min), 28 (ITS) or 32 (plastid loci) cycles of denaturation at 94°C (1 min), annealing 52°C (ITS) or 54°C (plastid loci) (1 min), elongation at 72°C (2 min), and a final elongation of 4 min. Amplified products were purified using precipitation with 11% solution of polyethylene glycol (PEG) 8000 and ethanol cleaning. Sequencing reactions in both directions were performed using BigDye Terminator 3.1 (Applied Biosystems, Carlsbad, California) chemistry and analyzed on an ABI3130XL sequencer (Applied Biosystems/Life Technologies Corporation, Carlsbad, California) following the manufacturer’s protocol at Universidade Estadual de Feira de Santana, Bahia, Brazil. Some PCR products were sequenced at the Interdisciplinary Center for Biotechnology Research at the University of Florida, Gainesville.

Sequence Assembly and Alignment—The sequences were edited using Geneious 6.1.8 (https://www.geneious.com) and aligned using the program Clustal2X (Larkin et al. 2007); alignments were checked by eye. Gaps were coded according to the "simple coding" criterion of Simmons and Ochoterena (2000) using the software Seqstate v.1.4.1 (Müller 2005).

The matrix includes 230 taxa and 2557 characters. DNA:1-2191 (nrETS = 1-635, nrITS = 636-2191); and standard (indels coded) 2192-2557.