The cichlid oral and pharyngeal jaws are evolutionarily and genetically coupled

Evolutionary constraints may significantly bias phenotypic change, while “breaking” from such constraints can lead to expanded ecological opportunity. Ray-finned fishes have broken functional constraints by developing two jaws (oral-pharyngeal), decoupling prey capture (oral jaw) from processing (pharyngeal jaw). It is hypothesized that the oral and pharyngeal jaws represent independent evolutionary modules and this facilitated diversification in feeding architectures. Here we test this hypothesis in African cichlids. Contrary to our expectation, we find integration between jaws at multiple evolutionary levels. Next, we document integration at the genetic level, and identify a candidate gene, smad7, within a pleiotropic locus for oral and pharyngeal jaw shape that exhibits correlated expression between the two tissues. Collectively, our data show that African cichlid evolutionary success has occurred within the context of a coupled jaw system, an attribute that may be driving adaptive evolution in this iconic group by facilitating rapid shifts between foraging habitats, providing an advantage in a stochastic environment such as the East African Rift-Valley.

almost all cases (Supplementary Data 17). To remove the allometric component of shape variation, we extracted the landmark residuals from this Procrustes ANOVA model to obtain landmark data sets for all four traits for use in subsequent analyses.
All structures were landmarked for both left and right sides separated by their midlines, allowing us to assess object symmetry and remove the effects of developmental noise. First, the landmarks from one side are reflected onto the other and undergo Procrustes superimposition.
The variation among the reflected and original landmark configurations for all comprises the symmetric component of shape variation 3 . The degree of symmetry was then statistically evaluated using Procrustes ANOVA using residual randomization permutation procedures 4 . All geometric morphometric datasets underwent symmetric corrections (Supplementary Data 18).

RNA Extraction and Quantitative PCR
We dissected tissues from three species of cichlid to extract RNA and perform quantitative PCR (qPCR): Labeotropheus fuelleborni (LF, n = 8), Maylandia callainos (MC, n = 8), and Tropheops kumara (TK, n = 6). We dissected lower oral jaw and lower pharyngeal jaw tissues, and took a caudal fin clip to act as a control tissue. All tissues were placed into TRIzol Reagent (Ambion Life Technologies) to limit RNA degradation and homogenized using the Bullet Blender Storm Tissue Homogenizer and stainless steel UFO beads (Next Advance, Averill Park, NY, USA). RNA was isolated from homogenized tissues via a phenol/chloroform extraction technique and ethanol precipitation. We removed any genomic DNA (gDNA) from the samples by degrading the gDNA using a DNase enzyme (Invitrogen). We quantified the amount of RNA in the samples spectrophotometrically (NanoDrop 2000, Thermo Scientific) and standardized across all samples to 100ng/μL. We reverse transcribed RNA to cDNA using a High Capacity cDNA Reverse Transcription Kit (Ambion Life Technologies).

4.
1. lower oral jaw, arrows denote changes in aspects of the jaw processes, particularly the attachment site for the A2 muscle on the ascending arm of the mandible. Changing the size and shape of this process will have consequences for jaw closing mechanics. c, PC1 lower pharyngeal jaw, arrows denote changes in overall width, length, and wing process size. Note the size of the wing processes where the ADD5 muscle attaches. Also, changing the overall length will impact the attachment site for the PH muscle on the pharyngeal jaw keel processes. d, PC2 lower pharyngeal jaw, arrows denote changes in depth of the jaw, which would likely impact the size of the attachment sites for the PC-I and PC-E muscles, and thus their contraction strength.