Brachiopods represent an animal phylum of benthic marine organisms that originated in the Cambrian. About 400 recent species are known from today's oceans (Emig et al., 2013). Around 5,000 fossil genera have been described, as brachiopods were dominant in the benthic marine environment during the Palaeozoic (Logan, 2007). Brachiopods have a biphasic life cycle with a planktonic larvae and sessile adults (Figure 1A). The phylum is divided into three clades namely Rhynchonelliformea and Craniiformea, which have short-lived lecithotrophic larvae and Linguliformea, which have long lived planktotrophic larva (Williams et al., 1996; Carlson, 2016).

Although various candidate stem-group brachiopods are known (Holmer et al., 2002, 2008, 2011; Balthasar, 2004; Skovsted et al., 2009a,b), no single hypothesis of early brachiopod body plan evolution yet commands a consensus, despite the potential of the Cambrian fossil record for reconstructing early body plan evolution in this, or any other, animal phylum (Budd and Jensen, 2000; Budd and Jackson, 2016). Thus, the early evolution of brachiopods is still a matter of debate and has led to the proposal of various scenarios with varying degrees of support.

One such scenario is the hypothesis of a “brachiopod fold,” which argues that brachiopods are transversely folded across the ontogenetic anterior-posterior axis (Figure 1B) (Cohen et al., 2003; Bitner and Cohen, 2013). According to this hypothesis, both valves are considered dorsal and in order to make useful comparisons with other animal phyla along the major body axis, brachiopods should be conceptually unfolded (Cohen et al., 2003). Since its original formulation, the brachiopod-fold hypothesis has gained support by some researchers in the brachiopod community with the suggestion that brachiopods arose by the folding of a Halkieria-like organism containing two protective shells at either end of the body (Benton and Harper, 2009). According to the brachiopod fold hypothesis, a folding process occurs during larval metamorphosis, as a rapid muscle mediated process that moves the posterior and anterior region of the larvae close together (Cohen et al., 2003). In this context, one hint about whether or not both valves can be considered dorsal would come from the analysis of gene expression patterns of developmental genes that are highly conserved among phyla. Such genes are ancient and can be traced to the last common ancestor of bilaterian animals (Schwaiger et al., 2014). If brachiopods evolved from a Halkieria-like organism by folding, one would expect the expression of genes that control the anterior and posterior domains in close proximity and opposed to each other.

Several studies have investigated the expression patterns of developmental genes in lecithotrophic brachiopod larvae (Altenburger et al., 2011; Santagata et al., 2012; Passamaneck et al., 2015; Martín-Durán et al., 2016; Vellutini and Hejnol, 2016). In these analyses it has been shown that during development the genes six3/6, NK2.1, gsc and otx are expressed in the anterior domain, which becomes the apical lobe in the rhynchonelliform Terebratalia transversa larva, and also in the anterior domain of the craniiform Novocrania anomala larva (Martín-Durán et al., 2016). Conversely, the genes evx and cdx (“posterior genes”) are expressed in the area that becomes the pedicle lobe and the posterior domain of the mantle lobe in T. transversa, and also in the posterior domain of the posterior lobe in N. anomala (Figure 1C) (Altenburger et al., 2011; Martín-Durán et al., 2016). Hox genes are not expressed collinearly in these brachiopod larvae (Schiemann et al., 2017). Analysis of the Hox cluster in T. transversa showed a split into three subclusters similar to that observed in other spiralians, such as in the annelid Capitella teleata and the limpet mollusc Lottia gigantea (Schiemann et al., 2017). Gene expression data for individuals during metamorphosis and for juveniles are still missing. Expression of Scr and Antp in the shell-forming epithelia of N. anomala and T. transversa larva suggests a role of these genes during juvenile shell formation (Schiemann et al., 2017).

The expression patterns of “anterior” and “posterior” genes in lecithotrophic brachiopod larvae are in an anterior-posterior sequence similar to the expression domains as detected in, for example, annelids and sea urchin embryos (Wei et al., 2012; Martín-Durán et al., 2016). As the morphogenetic events occurring during metamorphosis are known for T. transversa and N. anomala, it is possible to trace the body axes to the post-metamorphic body plan, and there are no signs of a folding event. Cohen et al. (2003) based the brachiopod fold hypothesis on observations during metamorphosis of N. anomala (Nielsen, 1991). However, a re-evaluation of metamorphosis in N. anomala showed that larva settle with the posterior-most tip of the posterior larval lobe and that vental and dorsal valves are not secreted from the same tissues (Altenburger et al., 2013).

Since there is no folding event during metamorphosis in craniiform or rhynchonelliform brachiopods (Altenburger and Wanninger, 2009; Altenburger et al., 2013), we can clearly state that there is no evidence in brachiopod ontogeny that supports the brachiopod fold hypothesis. Moreover, the only known exceptionally preserved lower Cambrian rhynchonelliform brachiopod Kutorgina chengjiangensis clearly has a straight gut (Zhang et al., 2007), indicating that the body axis orientation of brachiopods has been retained since the Cambrian.