Edible blueberry species are well recognized for their potential health benefits. Ericaceae fruits including the North American highbush blueberry (Vaccinium corymbosum L.) and five less common edible blueberry relatives from the New World tropics, Anthopterus wardii Ball, Cavendishia grandifolia Hoerld, Macleania coccoloboides A. C. Sm., Sphyrospermum buxifolium Poepp. & Endl., and Sphyrospermum cordifolium Benth, were investigated for their antioxidant properties and phenolic profiles. The Neotropical berries C. grandifolia and A. wardii exhibited significantly higher DPPH• and ABTS•+ scavenging and iron chelation activities than V. corymbosum. Total phenolic content and HPLC-PDA compositional fingerprint analyses were also carried out. Significant correlations were observed among total phenolic contents, DPPH• and ABTS•+ scavenging, and iron chelation activities. Using HPLC-PDA, the phenolic constituents in the berries were identified as chlorogenic acid, p-coumaric acid, hyperoside, quercetin-3-O-glucoside, isoorientin, isovitexin, orientin and vitexin. Principal components analysis reduced the dimensions of antioxidant and total phenolic data to two components, which accounted for 95% of total variation among the six fruits. Each fruit species formed its own cluster, and therefore the antioxidant profile of each species was shown to be distinct.

There are many neotropical blueberries, and recent studies have shown that some have even stronger antioxidant activity than the well-known edible North American blueberries. Antioxidant marker compounds were predicted by applying multivariate statistics to data from LC-TOF-MS analysis and antioxidant assays of 3 North American blueberry species (Vaccinium corymbosum, Vaccinium angustifolium, and a defined mixture of Vaccinium virgatum with V. corymbosum) and 12 neotropical blueberry species (Anthopterus wardii, Cavendishia grandifolia, Cavendishia isernii, Ceratostema silvicola, Disterigma rimbachii, Macleania coccoloboides, Macleania cordifolia, Macleania rupestris, Satyria boliviana, Sphyrospermum buxifolium, Sphyrospermum cordifolium, and Sphyrospermum ellipticum). Fourteen antioxidant markers were detected, and 12 of these, including 7 anthocyanins, 3 flavonols, 1 hydroxycinnamic acid, and 1 iridoid glycoside, were identified. This application of multivariate analysis to bioactivity and mass data can be used for identification of pharmacologically active natural products and may help to determine which neotropical blueberry species will be prioritized for agricultural development. Also, the compositional differences between North American and neotropical blueberries were determined by chemometric analysis, and 44 marker compounds including 16 anthocyanins, 15 flavonoids, 7 hydroxycinnamic acid derivatives, 5 triterpene glycosides, and 1 iridoid glycoside were identified.

Four flavone C-glycosides, isoorientin (1), orientin (2), vitexin (3), and isovitexin (4), were isolated from the neotropical blueberry of Anthopterus wardii, a so-called “superfruit”, using antioxidant activity-guided fractionation. A dose-response relationship of compounds 1–4 was determined for their anti-inflammatory activity against interleukin-8 (IL-8) and for the inhibition of matrix metalloproteinase-1 (MMP-1) expression, an inflammatory marker for chronic obstructive pulmonary disease. The four flavone C-glycosides exhibited inhibitory activity against IL-8 production and MMP-1 expression, with compounds 1, 3, and 4 having the most potent inhibitory activities in both assays at 100 μg/ml. The structures of compounds 1–4 were determined by spectroscopic methods. These flavone C-glycosides are reported for the first time in the Anthopterus genus.

Disterigma s. l. comprises 37 species of small shrubs found in Central and South America, mostly in cloud forests and páramos. The notion that the morphological circumscription of Disterigma s. l. is too wide and that it may not be a monophyletic group was supported by recent molecular evidence extracted from only a few species of Disterigma. This study addresses the monophyly and position of Disterigma s. l. within the complex blueberry tribe Vaccinieae through an intensive taxon sampling. Seventy-nine species are sampled, including 84% of the species of Disterigma s. l. and 13 genera of neotropical Vaccinieae. The parsimony analysis of nrITS and ndhF data reveal that Disterigma s. l. is polyphyletic with three independent origins within Vaccinieae. The majority of the species form a well supported monophyletic group, Disterigma s. s., which is located within a large Andean clade; all Disterigma s. s. species share the presence of a differentiated pair of apical bracteoles. Segregated from Disterigma s. s., but also Andean in origin, is the novel Central Andes Segregated Disterigma clade, where D. bracteatum, D. pentandrum, D. rimbachii, and D. ulei are found. Also resolved outside of Disterigma s. s. is D. trimerum, a taxon of Mesoamerican origin.

Disterigma sensu lato (s.l.) is a polyphyletic genus of neotropical blueberries that comprises 37 species found in Central and South America, mostly in Andean cloud forests and pa´ramos. The present phylogenetic analysis, based on molecular and morphological data for 84% of the species of Disterigma s.l., addressed the monophyly of the genus and identified the putative morphological synapomorphies of the main clades. The sensitivity of the phylogenetic hypotheses to the choice of analytical parameters was addressed in detail, by varying all possible parameters and applying different treatments for indel coding. A total of 29 transformation cost matrices were investigated and all these analyses were performed in POY 4, using direct optimization. A diagnosable Disterigma sensu stricto (s.s.) clade was recovered in all analyses, despite changes in the transformation costs. Additionally, none of the analyses refuted the segregation of D. trimerum, D. ulei, D. pentandrum, D. rimbachii, and D. bracteatum from Disterigma s.s. Although measurements of congruence are used to choose among alternative analytical parameter sets, the behaviour of the RILD index (rescaled incongruence length difference) in this study suggested that it may not be a good measure of congruence and that further studies are necessary to better understand its performance.