Constitutive gene expression differs in three brain regions important for cognition in neophobic and non-neophobic house sparrows (Passer domesticus)

Abstract

Neophobia (aversion to new objects, food, and environments) is a personality trait that affects the ability of wildlife to adapt to new challenges and opportunities. Despite the ubiquity and importance of this trait, the molecular mechanisms underlying repeatable individual differences in neophobia in wild animals are poorly understood. We evaluated wild-caught house sparrows (Passer domesticus) for neophobia in the lab using novel object tests. We then selected a subset of neophobic and non-neophobic individuals (n = 3 of each, all females) and extracted RNA from four brain regions involved in learning, memory, threat perception, and executive function: striatum, caudal dorsomedial hippocampus, medial ventral arcopallium, and caudolateral nidopallium (NCL). Our analysis of differentially expressed genes (DEGs) used 11,889 gene regions annotated in the house sparrow reference genome for which we had an average of 25.7 million mapped reads/sample. PERMANOVA identified significant effects of brain region, phenotype (neophobic vs. non-neophobic), and a brain region by phenotype interaction. Comparing neophobic and non-neophobic birds revealed constitutive differences in DEGs in three of the four brain regions examined: hippocampus (12% of the transcriptome significantly differentially expressed), striatum (4%) and NCL (3%). DEGs included important known neuroendocrine mediators of learning, memory, executive function, and anxiety behavior, including serotonin receptor 5A, dopamine receptors 1, 2 and 5 (downregulated in neophobic birds), and estrogen receptor beta (upregulated in neophobic birds). These results suggest that some of the behavioral differences between phenotypes may be due to underlying gene expression differences in the brain. The large number of DEGs in neophobic and non-neophobic birds also implies that there are major differences in neural function between the two phenotypes that could affect a wide variety of behavioral traits beyond neophobia.

Fig 1. Locations of brain punches used for house sparrow RNAseq.

Depiction of the approximate locations brain punches were taken from coronal 200 μm sections and the corresponding regions used as landmarks. Top (caudal): Medial ventral arcopallium (AMV) samples consisted of three 19 G punches and caudolateral nidopallium (NCL) samples consisted of two 15 G punches. The NCL was sampled on the following section after AMV, but the regions are pictured on the same slice for simplicity. Middle: Caudal dorsomedial hippocampus (HP) samples consisted of two 15 G punches. Bottom (rostral): Striatum (StM) samples consisted of two 11 G punches. Abbreviations: Cb = cerebellum, A = arcopallium, FA = tractus fronto-arcopallialis, LFS = lamina frontalis suprema, LPS = lamina pallio-subpallialis, COA = anterior commissure, OM = tractus occipito-mesencephalicus, TSM = tractus septopallio-mesencephalicus, QF = tractus quintofrontalis.

Fig 2

Top: Kaplan-Meier survival curves of house sparrow feeding likelihood in the presence of a novel object. There were four object trials for each bird, except for two missing trials (1 neophobic, 1 non-neophobic) where the video camera malfunctioned. Data are split by neophobia phenotype (not neophobic n = 8, neophobic n = 7) and with 95% confidence intervals. The trial ended at 3600 s. Bottom: Risk table indicating the number of sparrows yet to feed from the dish in 300 s intervals. Both plot and table were created using the ‘survminer’ package in R Studio [71].

Fig 3

A) Distance-based Redundancy Analysis (dbRDA) spider plot of gene expression, split by brain region and phenotype. Each brain region is represented by a different shape, and phenotypes are represented by colors (blue shades: not neophobic or “NotNeo”, n = 3 and red shades: neophobic or “Neo”, n = 3). Results from permutational multivariate analysis of variance (PERMANOVA) are shown. B) Venn diagram of genes differentially expressed between neophobic and not neophobic individuals. This diagram highlights the minimal overlap among different brain regions in the identities of differentially expressed genes.

Fig 4. Enriched eukaryotic orthologous group (KOG) terms in the house sparrow transcriptome across four brain regions.

Positive delta-ranks (red) are associated with upregulation in neophobic birds relative to non-neophobic birds, and significance is based on Benjamini-Hochberg adjusted p-values (FDR).