Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Comparative Study
. 2016 Jan 26;113(4):E469-78.
doi: 10.1073/pnas.1510903113. Epub 2016 Jan 6.

Transcriptional profiles of supragranular-enriched genes associate with corticocortical network architecture in the human brain

Fenna M Krienen  1 , B T Thomas Yeo  2 , Tian Ge  3 , Randy L Buckner  4 , Chet C Sherwood  5
Affiliations
Comparative Study

Transcriptional profiles of supragranular-enriched genes associate with corticocortical network architecture in the human brain

Fenna M Krienen et al. Proc Natl Acad Sci U S A. .

Abstract

The human brain is patterned with disproportionately large, distributed cerebral networks that connect multiple association zones in the frontal, temporal, and parietal lobes. The expansion of the cortical surface, along with the emergence of long-range connectivity networks, may be reflected in changes to the underlying molecular architecture. Using the Allen Institute's human brain transcriptional atlas, we demonstrate that genes particularly enriched in supragranular layers of the human cerebral cortex relative to mouse distinguish major cortical classes. The topography of transcriptional expression reflects large-scale brain network organization consistent with estimates from functional connectivity MRI and anatomical tracing in nonhuman primates. Microarray expression data for genes preferentially expressed in human upper layers (II/III), but enriched only in lower layers (V/VI) of mouse, were cross-correlated to identify molecular profiles across the cerebral cortex of postmortem human brains (n = 6). Unimodal sensory and motor zones have similar molecular profiles, despite being distributed across the cortical mantle. Sensory/motor profiles were anticorrelated with paralimbic and certain distributed association network profiles. Tests of alternative gene sets did not consistently distinguish sensory and motor regions from paralimbic and association regions: (i) genes enriched in supragranular layers in both humans and mice, (ii) genes cortically enriched in humans relative to nonhuman primates, (iii) genes related to connectivity in rodents, (iv) genes associated with human and mouse connectivity, and (v) 1,454 gene sets curated from known gene ontologies. Molecular innovations of upper cortical layers may be an important component in the evolution of long-range corticocortical projections.

Keywords: association cortex; brain evolution; corticocortical connectivity; human transcriptome; supragranular.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Fig. 1.
Fig. 1.
Transcriptional profiles follow cortical subtypes and network topography. (A) Correlations between pairs of 114 brain regions in terms of their transcriptional profiles for the HSE genes. Correlations are averaged across individuals (n = 6). Complete black rows are brain regions that were not sampled by any individual. (B) Correlation matrix showing coupling patterns across the same 114 regions, measured by fcMRI at rest [adapted from Buckner et al. (27)]. Regions in both matrices are arranged such that those that belong to the same fcMRI network are grouped together. (C, Left) Surface representation of the 17-network parcellation used to group brain regions in A and B [adapted from Yeo et al. (31)]. White asterisks are locations of regions shown in the polar plot shown on the right. (Right) Polar plot showing transcriptional profile correlations between the auditory cortex region and surrounding regions, as well as distant occipital regions. Note the higher similarity of transcriptional profiles between somato/motor regions and occipital visual regions than to neighboring association cortical regions.
Fig. S1.
Fig. S1.
Locations of included and excluded brain samples. (A) Approximate locations of brain samples on a Caret surface representation. Regions were included if centroid MNI coordinates as well as neighboring coordinates had the same network assignment. (B) This excluded subcortical structures and structures that lay on the borders between networks.
Fig. S2.
Fig. S2.
Transcriptional similarity within the default network. Polar plot showing transcriptional profile correlations between a medial prefrontal cortex region within the default network and other default network regions, as well as dorsal attention network regions. FEF, frontal eye fields; IPL, inferior parietal lobule; ParOcc, occipitoparietal cortex; PCC, posterior cingulate cortex; PFCd, dorsal prefrontal cortex; PFCm, medial prefrontal cortex; PostC, postcentral gyrus; Temp, temporal cortex; SPL, superior parietal lobule; TempOcc, occipitotemporal cortex.
Fig. S3.
Fig. S3.
Significant edges in the HSE set correlation matrix and in the difference in correlation strengths between HSE and alternative gene sets. For each heatmap plot, network-averaged correlation coefficients for the 17 networks are color coded as in Fig. 1 A and C. For each binary plot, a thresholded matrix shows network pairs in the correlation matrix that are significant at FDR-corrected q < 0.05 in white. (A) Significant edges in the HSE correlation matrix are determined by permuting gene labels between pairs of networks. Average expression for each of the 19 genes is computed across all brain samples falling within a given network before the correlation is computed. (B) Significantly different edges between the HSE set and comparison sets are determined by permuting genes between the two sets for each brain sample and averaging correlations for regions that fall within the same network. See text in Supporting Information for additional details.
Fig. S4.
Fig. S4.
Spearman’s ρ is robust against outliers. (A) Network-averaged correlation coefficients for the 17 networks for the HSE genes using either Pearson’s product-moment correlation or rank-signed Spearman’s ρ. Resulting matrices are comparable using the two metrics. (B) Correlation matrices for one of 13 MSigDB gene sets that had large connected components using the NBS metric (see main text). Although weaker than the HSE set, each set showed strong within-network similarity as well as anti-correlations between sensory and association/paralimbic networks. Inspection of these sets revealed one gene in common: CARTPT. For 11 of 13 gene sets, CARTPT appeared to be an outlier, as removing it from these sets substantially diminished the within- and across-network correlations using Pearson’s r. Using Spearman’s ρ attenuates the influence of such outliers. (C, Left) Network-averaged correlation coefficients for MSigDB Biological Process Set #456 (Regulated Secretory Pathway), averaged into the 17 networks. (Right) MSigDB Biological Process Set #660 (Neurotransmitter Secretion). The two sets are highly overlapping; the Neurotransmitter Secretion set is a subset of 13 of 15 genes contained in the Regulated Secretory Pathway set. Both sets contain CARTPT. Unlike other MSigDB gene sets (e.g., B), the correlation structure persists when this gene is removed.
Fig. 2.
Fig. 2.
Individual subjects show consistent transcriptional profile groupings across networks. Individual correlation matrices for the HSE genes. Rows and columns are arranged by network, as in Fig. 1. Strong positive correlations along the diagonal indicate that transcriptional profiles are similar in brain regions that belong to the same network. Strong positive correlations on the off-diagonal indicate similarity between networks.
Fig. 3.
Fig. 3.
Transcriptional profiles are more similar within-network than between networks. (Left) HSE set transcriptional similarity (correlation) summarized by the seven-network parcellation. Regions that belong to the same network have higher correlations than regions that belong to different networks. Error bars show SEM across individuals. (Right) Surface representation of plotted networks.
Fig. 4.
Fig. 4.
Relative differences in transcriptional profiles across networks are revealed by detrending expression values. (Left) Mean expression values for each of the 19 genes in the HSE gene set, plotted for seven networks. (Right) After mean expression is subtracted for each probe across cortical samples, relative differences in expression become evident across the seven networks.
Fig. 5.
Fig. 5.
Transcriptional profiles can be similar even when located far apart on the cortical mantle. Two individual matrices from Fig. 2 are rearranged according to rostrocaudal position on the cortical mantle. In each case, strong off-diagonal correlations indicate that regions spaced far apart have similar transcriptional profiles.
Fig. 6.
Fig. 6.
HSE gene set transcriptional profiles are a function of spatial proximity as well as network identity and cortical type. Transcriptional profile correlations are plotted against Euclidean distance measurements for pairs of brain samples. Dark gray points are within-network pairs. Note that dark gray points tend to group at the top of the graph, meaning they have higher correlations even at long distances. Red points are somato/motor network to visual network pairs. Note that they tend to have high correlations despite long distances. Blue points are paralimbic or ventral attention to default network pairs. Note that they tend to have high correlations at both close and long distances. Magenta points are paralimbic or ventral attention to visual network or somato/motor network pairs. Note that they tend to have low correlations at all distances. Light gray points are all other combinations. Note that they tend to follow the overall negative correlation and are particularly evident at the long-range, low-correlation corner of the graph.
Fig. S5.
Fig. S5.
HSE gene set transcriptional profiles are a function of spatial proximity as well as network identity and cortical type. Transcriptional profile correlations are plotted against Euclidean distance measurements for pairs of brain samples. Dark gray points are within-network pairs. Notice that dark gray points tend to group at the top of the graph, meaning that they have higher correlations even at long distances. Red points are visual network to somato/motor network pairs. Note that they tend to have high correlations despite long distances. Blue points are paralimbic or ventral attention to default network pairs. Note that they tend to have high correlations at both close and long distances. Magenta points are paralimbic or ventral attention to visual network or somato/motor network pairs. Note they tend to have low correlations at all distances. Light gray points are all other combinations. Note that they tend to follow the overall negative correlation and are particularly evident at the long-range, low-correlation corner of the graph.
Fig. S6.
Fig. S6.
Effect of spatial proximity and network identity on alternative gene sets. (A) Rodent Connectivity gene set. (B) Conserved Supragranular gene set. Color-coding follows Fig. S5.
Fig. S7.
Fig. S7.
Effect of spatial proximity and network identity on alternative gene sets. (A) Human Cortically Enriched gene set. (B) Human/Mouse Connectivity gene set. Color-coding follows Fig. S5.
Fig. 7.
Fig. 7.
Balance of local and distant coupling predicts transcriptional similarity. (A) Relative difference in distant–local degree connectivity measured by proportion of correlations that are within the local neighborhood versus distant correlations in resting state functional connectivity data [adapted from Sepulcre et al. (13)]. (B) Average distance score for 17 networks, listed along the group-averaged transcriptional similarity matrix for the HSE gene set. (C) Scatter plot showing the relationship between transcriptional similarity and degree connectivity score (from A) for all network pairs.
See this image and copyright information in PMC

References

    1. Khaitovich P, Enard W, Lachmann M, Pääbo S. Evolution of primate gene expression. Nat Rev Genet. 2006;7(9):693–702. - PubMed
    1. Konopka G, et al. Human-specific transcriptional networks in the brain. Neuron. 2012;75(4):601–617. - PMC - PubMed
    1. Oldham MC, et al. Functional organization of the transcriptome in human brain. Nat Neurosci. 2008;11(11):1271–1282. - PMC - PubMed
    1. Lein ES, et al. Genome-wide atlas of gene expression in the adult mouse brain. Nature. 2007;445(7124):168–176. - PubMed
    1. Babbitt CC, et al. Both noncoding and protein-coding RNAs contribute to gene expression evolution in the primate brain. Genome Biol Evol. 2010;2(0):67–79. - PMC - PubMed

Publication types

MeSH terms

LinkOut - more resources