"An RNA gene expressed during cortical development evolved rapidly in humans".

Katherine S. Pollard ^{1, 8, 9}, Sofie R. Salama ^{1, 2, 9}, Nelle Lambert ^{4, 5}, Marie-Alexandra Lambot ⁴, Sandra Coppens ⁴, Jakob S. Pedersen ¹, Sol Katzman ¹, Bryan King ^{1, 2}, Courtney Onodera ¹, Adam Siepel ^{1, 7}, Andrew D. Kern ¹, Colette Dehay ^{6, 7}, Haller Igel ³, Manuel Ares, Jr ³, Pierre Vanderhaeghen ⁴, and David Haussler ^{1, 2}

¹ Center for Biomolecular Science & Engineering,
² Howard Hughes Medical Institute, and
³ Center for Molecular Biology of RNA, Department of Molecular, Cell & Developmental Biology, University of California, Santa Cruz, California 95064, USA
⁴ Institut de Recherche Interdisciplinaire en Biologie Humaine et Moléculaire (IRIBHM), Free University of Brussels (ULB), and
⁵ Department of Psychiatry, Erasme Hospital, Free University of Brussels (ULB), B-1070 Brussels, Belgium
⁶ INSERM, U371, Stem Cell & Brain Research Institute, F-69500 Bron cedex, France
⁷ Université Claude Bernard, Lyon 1, F-69500 Bron cedex, France
⁸ †Present addresses: Department of Statistics and UC Davis Genome Center, University of California, Davis, California 95616, USA (K.S.P.); Department of Biological Statistics and Computational Biology, Cornell University, Ithaca, New York 14853, USA (A.S.)
⁹ *These authors contributed equally to this work

Correspondence to: David Haussler ^{1, 2} (Email: haussler@soe.ucsc.edu).

The sequences of the HAR1 transcripts reported in this study have been submitted to GenBank (accession numbers DQ860409–DQ860415).

The developmental and evolutionary mechanisms behind the emergence of human-specific brain features remain largely unknown. However, the recent ability to compare our genome to that of our closest relative, the chimpanzee, provides new avenues to link genetic and phenotypic changes in the evolution of the human brain. We devised a ranking of regions in the human genome that show significant evolutionary acceleration. Here we report that the most dramatic of these 'human accelerated regions', HAR1, is part of a novel RNA gene (HAR1F) that is expressed specifically in Cajal–Retzius neurons in the developing human neocortex from 7 to 19 gestational weeks, a crucial period for cortical neuron specification and migration. HAR1F is co-expressed with reelin, a product of Cajal–Retzius neurons that is of fundamental importance in specifying the six-layer structure of the human cortex. HAR1 and the other human accelerated regions provide new candidates in the search for uniquely human biology.

The hallmark of evolutionary shift of function is sudden change in a region of the genome that previously has been highly conserved owing to negative selection. It has been speculated that changes of this type in FOXP2 [1], a gene involved in speech production, and ASPM [2], which affects brain size, have had a significant role in the evolution of the human brain (reviewed in ref. 3). The vast majority of the approximately 15 million changes in our genome since our common ancestor with the chimpanzee are likely to represent neutral drift [4,5], so systematic searches for potentially important evolutionary acceleration have focused exclusively on protein coding regions [5,6,7,8,9], where there is a more favourable signal-to-noise ratio. However, protein coding regions account for only about one-third of the segments in the human genome thought to be under negative selection [10], and thus these searches may be missing the majority of the functional elements in the genome [5,11]. With the availability of nearly complete genome sequences for several vertebrates, comparative genomics can now be used to predict functional elements in the 98.5% of the genome that is non-coding, through patterns of ancestral negative selection [10,12,13,14,15,16]. Here we scan these ancestrally conserved genomic regions to find those that show a significantly accelerated rate of substitution in the human lineage since divergence from our common ancestor with the chimpanzee. Many of the human accelerated regions (HARs) found in this scan are associated with genes known to be involved in transcriptional regulation and neurodevelopment. HAR1, the most dramatically changed element, is part of a novel RNA gene expressed during human cortical development.

Identification of human accelerated regions

To overcome the very low signal-to-noise ratio among changes in non-coding regions, we first searched independently for maximal-length regions of the chimpanzee genome with at least 96% identity over 100 base pairs (bp) with the orthologous regions in mouse and rat, suggesting a significant level of negative selection in these regions. For each of the approximately 35,000 such mammalian conserved regions (median length, 140 bp), we then examined the orthologous segments in all other available amniote genomes, looking for regions that have a large number of non-adjacent changes in human relative to other species. A likelihood ratio test was used to rank and evaluate the significance of each region. Controlling the genome-wide false discovery rate (FDR) to be less than 5%, we identified 49 regions with a statistically significant substitution rate increase in human. We compared the results of this analysis to those from several complementary approaches (see Supplementary Notes S1) and found that the fastest-evolving regions score uniformly high regardless of the method (Supplementary Notes S2). Of the 49 HARs, 96% are in non-coding segments (Supplementary Table S2). Gene Ontology terms related to DNA binding and transcriptional regulation are significantly enriched among the genes adjacent to the HARs (Supplementary Table S8), and 24% of the HARs are adjacent to a neurodevelopmental gene (Supplementary Table S7), making these regions especially interesting as candidates for brain-specific regulatory elements that may have changed significantly during human evolution.

HAR1 lies in a pair of novel non-coding RNA genes

The 118-bp HAR1 region showed the most dramatically accelerated change (FDR-adjusted P < 0.0005), with an estimated 18 substitutions in the human lineage since the human–chimpanzee ancestor, compared with the expected 0.27 substitutions on the basis of the slow rate of change in this region in other amniotes (Supplementary Notes S3). Only two bases (out of 118) are changed between chimpanzee and chicken, indicating that the region was present and functional in our ancestor at least 310 million years (Myr) ago. No orthologue of HAR1 was detected in the frog (Xenopus tropicalis), any of the available fish genomes (zebrafish, Takifugu and Tetraodon), or in any invertebrate lineage, indicating that it originated no more than about 400 Myr ago [17]. No paralogues were detected in any amniote genome draft. Resequencing in four primates further confirms that all 18 substitutions are very likely to have occurred in the human lineage (Supplementary Fig. S2). Checking against available single nucleotide polymorphisms (SNPs; http://www.ncbi.nlm.nih.gov/SNP/) [18] and resequencing this region in a 24-person diversity panel [19] indicates that all 18 substitutions are fixed in the human population. Evidence from preliminary resequencing of a 6-kilobase (kb) region containing HAR1 shows levels of polymorphism and a positive skew in the frequency spectrum that are typical of European samples [20], suggesting that a recent selective sweep in this region of the genome is unlikely (Supplementary Notes S4). Thus, the changes in HAR1 clearly occurred on the human lineage, but probably took place more than 1 Myr ago.

HAR1 lies in the last band of chromosome 20q and is part of a pair of overlapping divergently transcribed genes, HAR1F and HAR1R, supported by expressed sequence tag (EST) evidence. Analysis by polymerase chain reaction with reverse transcription (RT–PCR), and 5' and 3' rapid amplification of complementary DNA ends (RACE), revealed a two-exon HAR1F gene and an alternatively spliced HAR1R gene with isoforms of two and three exons (Fig. 1, Supplementary Notes S5). We sequenced the entire HAR1F genomic region in chimpanzee, gorilla, orang-utan and crab-eating macaque (Supplementary Notes S4, Supplementary Fig. S3). The very high divergence of the non-human primates from the human reference found in HAR1 extends through most of the first HAR1F exon. In addition to the primates, the dog and cow genomes also contain HAR1F orthologues, whereas only the HAR1 region, and not the entire transcript, can be aligned to other available amniote genomes. Most HAR1R regions are highly divergent, except in the chimpanzee and macaque genomes.

Figure 1: HAR1-associated transcripts in genomic context.

Figure 1 : HAR1-associated transcripts in genomic context.

Schematic drawing showing the genomic context on chromosome 20q13.33 of the HAR1-associated transcripts HAR1F and HAR1R (black, with a chevroned line indicating introns), and the predicted RNA structure (green) Fig. 2, based on the May 2004 human assembly in the UCSC Genome Browser [41]. The level of conservation in the orthologous region in other vertebrate species (blue) is plotted for this region using the PhastCons program [16]. Both the common and testes-specific splice sites are conserved (data not shown).

We have also cloned HAR1-containing transcripts from mice (Supplementary Notes S6). As in human, we found overlapping transcripts that contain HAR1. As expected by the poor conservation in the surrounding genomic region, the homology between the human and mouse HAR1 transcripts is limited to an approximately 180-bp region surrounding the HAR1 segment. Thus, although the region of overlap between human HAR1F and HAR1R (including the HAR1 segment) was conserved from our common ancestor with chicken to our common ancestor with chimpanzee, the longer HAR1F and HAR1R transcripts have evolved much more rapidly during this period.

The HAR1F RNA forms a stable structure

Protein coding potential of the mouse and human HAR1F and HAR1R RNAs is poor or absent (Supplementary Notes S7). However, a phylogenetic stochastic context-free grammar model for RNA evolution [12] gives a strong prediction that the HAR1 region of the HAR1F transcript can fold into a stable RNA structure (Fig. 2), and a slightly less-strong prediction for the same region of the HAR1R transcript (data not shown). Three of five RNA helices in the predicted secondary structure are statistically confirmed by five pairs of human-specific compensatory changes that preserve Watson–Crick pairing: four pairs of transitions and one pair of transversions (P = 3.95 times 10^-8; Supplementary Notes S8). A different compensatory pair of transversions is also found in the platypus genome. No changes inconsistent with the core of any helix, excluding the terminal base pairs and the second pair of the outermost enclosing helix in mouse, were found in any sequence match from the National Center for Biotechnology Information (NCBI) trace repository (one match from each of 11 amniote species; Supplementary Fig. S2).

Figure 2: Predicted RNA secondary structure for HAR1F.

Figure 2 : Predicted RNA secondary structure for HAR1F.

a, Section of the multiple alignment of HAR1F in various amniote species. The multiple alignment is annotated with the secondary structure (fold) shown in panel b. Matching round parentheses indicates pairing bases. Square parentheses are used to indicate bases that are predicted to pair outside the region shown. Unpaired regions are shown in grey and substitutions are shown with the following colour scheme: green denotes compensatory transitions, yellow–green denotes compensatory transversions, purple denotes substitutions in unpaired regions, and red denotes non-compensated changes.

b, The evolutionarily conserved parts of the RNA secondary structure of the HAR1F region as predicted using the EvoFold program [12]. Substitutions are shown using the colour scheme in panel a; red bars indicate the region for which the alignment is shown in panel a. The structure is supported by substitutions on the human lineage, as well as a pair of changes in platypus (indicated by (P)). Only the helices with the compensatory substitutions can be considered to be supported by evolutionary data.

c, Distinct secondary structure predictions for helices C and D in the human sequence (Human) versus non-human sequences (Chimp), on the basis of combined comparative sequence analysis and structure probing. Circles indicate the bases that can be modified by DMS and indicate strong (black), weak (grey) or no (white) modification as shown in the gel to the right. For each species, lanes 1–4 show the sequencing reactions labelled according to the corresponding base in the RNA structure; lanes 5–8 show the primer extension reactions of RNA after no (- ), 0, 10 or 30 min of DMS treatment at 18 °C. Numbers on the right indicate the position in RNA structure; arrows indicate the modified base at position 61; solid lines indicate the position of bases 37–39; and dashed lines indicate the position of bases 54–57. The presence of a band in the last two lanes indicates modification of the base at the previous position (that is, the arrows at position 62 on the gel indicate a modified A at position 61).

We tested this secondary structure prediction by performing structure probing experiments using dimethyl sulphate (DMS) treatment and primer extension of in vitro synthesized human and chimpanzee HAR1F RNAs (Supplementary Notes S9) [21]. The results were largely consistent with the predicted structure, but we observed several differences between the human and chimpanzee RNAs (Fig. 2c and Supplementary Fig. S10). In particular, our results are consistent with a human-specific structure where helix D is extended at the expense of helix C (compare Fig. 2b to Fig. 2c). In contrast, excluding the human sequence from the comparative analysis leads to an evolutionarily conserved structure with three additional base pairs in helix C and two fewer pairs in helix D. This structure agrees with the protection of bases in helix C and the accessibility of bases in the shorter helix D we observe for the chimpanzee RNA. In the human RNA, bases in helix D are strongly protected whereas those in helix C are more accessible by DMS. These results indicate that although selective forces to preserve base pairing interactions shaped this region, the structure of the HAR1F RNA may be significantly different between humans and their ancestors, and may even lack helices C and E. This RNA structure is novel; no primary or secondary structural homology was found with any RNA in the Rfam RNA repository [22], and no convincing evidence of microRNA (miRNA) precursors in these transcripts has yet been found (Supplementary Notes S10). On the basis of this analysis, it seems that HAR1F and HAR1R are novel non-coding RNA genes.

HAR1F is expressed in the developing human and primate brain.

To determine if HAR1F is expressed in the developing human central nervous system, we performed RNA in situ hybridizations on human embryonic brain sections. We found that HAR1F is strongly and specifically expressed in the developing neocortex early in human embryonic development (Fig. 3). HAR1F was first detected at 7 and 9 gestational weeks (GW), specifically in the dorsal telencephalon (the primordium of the cerebral cortex) and not detected in other parts of the forebrain (Fig. 3a and data not shown). Within the early developing cortex, HAR1F was selectively expressed in a subset of cells located close to the pial surface in the marginal zone (Fig. 3a, b), indicating its expression in Cajal–Retzius neurons. This was confirmed by co-detection of reelin protein, a specific marker of these neurons, and HAR1F RNA (Fig. 3a, c, d), which showed that both HAR1F and reelin are expressed in the same cells. In contrast, HAR1F did not seem to be expressed in interneurons expressing vesicular GABA (gamma-aminobutyric acid) transporter (vGAT), the other main neuronal population of the marginal zone (Fig. 3d). The expression of HAR1 in reelin-positive neurons was maintained until at least 17–19 GW (Fig. 3c, d), at which stage reelin-positive Cajal–Retzius-like neurons are interspersed within the subpial granular layer (SGL)—a layer thought to contain Cajal–Retzius neurons and their precursors migrating tangentially from outside the neocortex [23,24,25], and which is uniquely developed in humans [24,25,26,27]. HAR1F was also expressed in cells of the upper cortical plate (Fig. 3b, c), presumably corresponding to neurons finishing their radial migration. Notably, the human reelin receptor gene, VLDLR, also shares this expression pattern in humans, but is not expressed in Cajal–Retzius neurons in the mouse [28]. At later stages (24 GW), expression of HAR1F was no longer detectable in Cajal–Retzius cells. At 17–24 GW, expression of HAR1F was observed in brain regions other than the cortex: in particular, the hippocampal primordium, the dentate gyrus, the developing cerebellar cortex, and a few hindbrain nuclei such as the olivar complex (Fig. 4). In contrast to the distinct pattern of HAR1F expression, probes specific for HAR1R show no specific staining in these embryonic sections (data not shown).

Figure 3: Expression of HAR1F and HAR1R in the developing neocortex.

Figure 3 : Expression of HAR1F and HAR1R in the developing neocortex.

Pattern of expression of HAR1F and reelin in the human and macaque developing brain, revealed by RNA in situ hybridization (HAR1F) and immunohistochemistry (reelin).

a, In situ hybridization of HAR1F antisense (AS) and sense (S) probes on 9-GW human brain coronal sections, illustrating selective expression of HAR1F in the dorsal telencephalon (DT), whereas the ventral telencephalon (VT) and thalamus (Th) show no obvious expression. The HAR1 probes contain 112 bp of the region of overlap between HAR1F and HAR1R. Therefore, the HAR1F sense probe has the potential to detect HAR1R expression. Within the DT, HAR1F is selectively enriched in the most superficial part of the cortex, much like reelin at the same stage (arrows). Scale bars: 500 microm on left panels; 250 microm on right panels. Dorsal (D) is top; lateral (L) is right on left panels.

b, At 11 GW, HAR1F expression is confined to the cerebral cortex, with expression in large cells situated at the border of the marginal zone (MZ) (arrows), corresponding to Cajal–Retzius neurons, and in the upper cortical plate (CP) (arrowheads). Scale bar: 125 microm for the right panel. IZ, intermediate zone; SVZ, subventricular zone; VZ, ventricular zone.

c, At 17 GW, expression of HAR1F and reelin remain confined to Cajal–Retzius cells in the MZ (arrows), intermingled with other cells of the subpial granular layer (SGL). In addition, HAR1F is expressed in neurons of the upper CP (arrowheads). Scale bar: 1,000 microm.

d, At 19 GW, HAR1F and reelin show the same pattern in the MZ/SGL (arrows), and HAR1F remains expressed in the upper CP (arrowheads), whereas the HAR1F sense probe shows no signal. Lower panels centred on the MZ illustrate double in situ hybridization/immunohistochemistry experiments. Lower left and middle panels show that HAR1F (in dark blue) and reelin (in dark brown) are co-localized in most cells, corresponding to Cajal–Retzius neurons (black arrows), although some reelin-positive cells do not express HAR1F (brown arrows). The lower right panel illustrates that HAR1F cells (blue arrows) seem to be mainly distinct from interneurons (brown arrows) expressing vesicular GABA transporter (vGAT). Scale bar: 250 microm.

e, Expression of HAR1F in macaque developing cortex. In situ hybridization (HAR1F) and immunohistochemistry (reelin) on E79 macaque brain parasagittal sections. As in the human, HAR1F and reelin are expressed in Cajal–Retzius cells in the MZ (arrows), intermingled with other cells of the SGL. In addition, HAR1F is expressed in neurons of the upper CP (arrowheads). Scale bar: 1,000 microm. The HAR1F sense probe does not reveal any obvious expression pattern.

Figure 4 : Expression of HAR1F in other parts of the developing brain.

Pattern of expression of HAR1F in the developing brain, revealed by RNA in situ hybridization.

a, In situ hybridization on 19-GW coronal brain sections, illustrating expression of HAR1F in the hippocampal primordium (arrows) and dentate gyrus (arrowheads). HAR1F sense probe detects no obvious signal. Scale bar: 750 microm

b, In situ hybridization on 24-GW coronal brain sections, illustrating expression of HAR1F in the cerebellar cortex (arrows in left panel) and olivar complex (arrows in right panel).

We examined the evolutionary conservation of this expression pattern by performing RNA in situ hybridizations on brain sections of another primate, the cynomolgus macaque. For these experiments, we cloned a fragment of macaque DNA corresponding to most of HAR1F exon 1 and tested it on embryonic day (E) 79–85 samples, corresponding to midgestation and to stages of active cortical neuron production and radial migration [29]. The HAR1F expression pattern was very similar to that observed in the human sections (Fig. 3e). HAR1F is co-expressed with reelin in the SGL and is also expressed in cells of the upper cortical plate. Thus, despite extensive sequence changes, the expression pattern of HAR1F in the developing cortex has been highly conserved since the divergence of hominoids and Old World monkeys some 25 Myr ago. This suggests that the HAR1F expression pattern is functionally significant and is reminiscent of the evolution of ASPM, which shows rapid change in its coding sequence but conserved patterns of expression in neural progenitors [2, 30].

HAR1 genes are expressed in the adult brain

We also examined HAR1F and HAR1R expression in adult human samples by using tissue RNA preparations for northern blot and quantitative real-time PCR (qPCR) analysis (Table 1), as well as RNA in situ hybridization. In situ hybridization experiments on adult brain sections revealed a diffuse pattern of staining for HAR1F in the frontal cortex and hippocampus (Supplementary Fig. S12). Although no specific signal was detected for either transcript by northern blot analysis (data not shown), we detected expression of both transcripts by qPCR in total brain, ovary and testes samples. Neither transcript was detected in other tissues tested, including adrenal, bladder, breast, colon, liver, pancreas, placenta, skeletal muscle and thymus. To further characterize the expression in brain, we obtained RNA samples from specific adult human brain regions as well as a fetal brain RNA sample. Expression of HAR1F is highest in cerebellum and is also prominent in forebrain structures, including the cortex, hippocampus, thalamus and hypothalamus. In general, HAR1F expression is higher than HAR1R. Testis is an exception; here transcripts are present in small, but roughly equal, amounts. It is interesting to note that HAR1R levels are dramatically lower than HAR1F (50-fold lower) in the fetal brain sample, consistent with our inability to detect it with in situ hybridization. The finding that both transcripts overlap and are expressed in the same tissues raises the possibility of antisense regulation between them. So far, however, we have not detected expression of the two transcripts in the same cells.

Table 1: Expression of human HAR1F and HAR1R.

Table 1 - Expression of human HAR1F and HAR1R.

In a qPCR analysis of mouse HAR1F and HAR1R expression (Supplementary Notes S6), we found that both HAR1F and HAR1R are expressed in embryonic development, and the peak expression of both HAR1F and HAR1R was at E15. In contrast to our observations in human fetal brain samples, HAR1F expression was not significantly higher than HAR1R expression. In adult tissues, we found both HAR1F and HAR1R specifically expressed in brain regions, but again the levels of HAR1F and HAR1R were similar. These results suggest that the regulation of the HAR1 region transcripts may have changed since the common ancestor of mouse and human.

HAR1 substitutions show a weak-to-strong bias

A curious feature of the evolution of HAR1 is that all 18 human-specific changes in the predicted RNA structure are from A or T ('W') nucleotides to G or C ('S') nucleotides. This has the effect of strengthening RNA helices in the predicted RNA structure. However, this pattern of highly biased substitutions extends well beyond the predicted RNA structure in a 1.2-kb region overlapping the first exon of HAR1F (34 W right arrow S substitutions out of 44 total human–chimpanzee differences, with no S right arrow W substitutions). Notably, a W right arrow S substitution bias is present in many of the human accelerated regions. Among the 49 most accelerated, the frequency of W right arrow S substitutions is 77% higher than that of S right arrow W substitutions, even after a correction for variation in the overall base composition of the ancestral sequence, whereas there is no difference in the larger set of 35,000 conserved regions (Supplementary Notes S11). Compared with an expected <5%, we find that 12% of these 49 regions, including HAR1, lie in the final band of their chromosome arms, where there is an increased rate of recombination [14]. Recombination hotspots are known to differ in location between human and chimpanzee [31], and have been speculated to be associated with increases in W right arrow S substitutions through a process known as biased gene conversion [32]. Selection for increased gene stability or increased expression levels have been proposed as alternative explanations for W right arrow S substitution biases [33,34]. Bias towards W right arrow S substitution through any of these general processes may have combined with selection for specific compensatory changes in the RNA structure to dramatically alter HAR1 during human evolution.

Conclusions

One primary issue remaining concerns the role of HAR1F and HAR1R in cortical development. Although HAR1F is co-expressed with reelin—a known regulator of cortical development—within Cajal–Retzius cells of the SGL, it remains to be determined whether HAR1F influences the expression of reelin or its receptors either directly or indirectly, or whether it acts through a different pathway. Notably, the extra-cortical sites of expression of HAR1F (such as cerebellar cortex and olivar nuclei) also correspond to structures that are patterned by the reelin pathway in mouse and human [35]. A suggested association between reelin-expression-related defects in the brain and schizophrenia provides additional motivation for further investigation of this issue [36]. Additionally, the pattern of expression of HAR1R suggests, but does not prove, that it may be expressed later in development to downregulate HAR1F by antisense inhibition, as is the case for some other developmental genes [37].

Our comparative genomic approach to identify the most dramatically changed segments of DNA in the human genome has identified a number of new candidate regions to test for clues in the attempt to decode the key events in human evolution. The first on this list seems to define a new type of RNA molecule expressed at a very crucial time and place in the development of the neocortex. Many of the other top candidates are associated with genes known to be involved in neurodevelopment, an area where there has been significant divergence since our last common ancestor with chimpanzee. Thus, this seems to be a promising approach to identifying candidate regions involved in neurodevelopmental aspects of our uniquely human biology.

Methods

Detailed methods are given in Supplementary Information. Using alignments produced by MULTIZ (http://www.bx.psu.edu/miller_lab/), we identified 34,498 conserved regions of the chimpanzee genome that are greater than or equal to100 bp long and greater than or equal to96% identical with mouse and rat. Each conserved region was evaluated for acceleration in the human lineage using a likelihood ratio test (LRT) using DNA from other species. Statistical significance was assessed by simulation and P-values are adjusted for multiple comparisons.

For RT–PCR, RACE and qPCR analysis, RNAs (Ambion and Clontech) were used to synthesize cDNA primed with a combination of oligo-dT and random hexamer primers using Superscript III (Invitrogen) according to the manufacturer's recommendations.

The qPCR analysis used iQSYBR Green Supermix (BioRad) with 2.5 ng microl-1 (or 0.8 ng microl-1 for polyA+-derived samples) cDNA template in an i-Cycler thermal cycler (BioRad); RACE analysis used the GeneRacer kit (Invitrogen). Secondary structure models were based on an EvoFold prediction [12], and chemical probing of unpaired A and C residues with dimethyl sulphate (DMS) was carried out using methods described previously [21].

RNA in situ hybridization using digoxigenin-labelled RNA probes adapted to human brain specimens was performed as described previously [38]. Immunodetection was performed as described [39] using mouse monoclonal anti-reelin antibody 142 (ref. 40; kindly provided by A. Goffinet) and a rabbit antibody to vGAT (Synaptic Systems). Between 2 and 16 sections from each case were tested.

This study was approved by the three relevant local ethics committees (Erasme Academic Hospital, University of Brussels, and Belgian National Fund for Scientific Research) on research involving human subjects. Written informed consent was given by the parents in each case. All experiments were performed in compliance with the national and European laws, as well as with INSERM institutional guidelines, concerning animal experimentation.

Acknowledgements

We thank G. Bejerano, C. Lowe, J. Kent, H. Noller, D. Feldheim, A. Love (UCSC), V. Albert and J.-C. Noel (Erasme Hospital), J.-P. Brion (ULB), A. Goffinet (UCL Louvain), the UCSC Genome Browser Group, Webb Miller (Penn. State), the Macaque Genome Sequencing Consortium, and the Broad Institute Genome Sequencing Platform and Whole-Genome Assembly Team. This work was funded by the Howard Hughes Medical Institute (D.H., S.R.S. and B.K.), the US NHGRI (D.H., A.D.K., S.K. and C.O.), the US National Cancer Institute (J.S.P.), the US NIGMS (K.S.P. and M.A.), the University of California Biotechnology Research and Education Program (A.S.), the Danish Research Council (J.S.P.), the Belgian FNRS, the Belgian FRSM, the Belgian Queen Elizabeth Medical Foundation (FMRE), the Interuniversity Attraction Poles Programme, Belgian State and Federal Office for Scientific, Technical and Cultural Affairs (P.V.), and the Fondation Erasme (M.-A.L. and N.L.). P.V. and M.-A.L. are Research Associate and Research Fellow of the FNRS, respectively. Author Contributions Computational methods developed and applied by K.S.P., J.S.P., A.S. and A.D.K. Experimental analysis by S.R.S., P.V., N.L., M.-A.L., S.C., S.K., B.K., C.O., C.D., H.I. and M.A. Manuscript primarily written by K.S.P., S.R.S., P.V. and D.H.
Competing interests statement:

The authors declared no competing interests.

Supplementary information accompanies this paper:
http://www.nature.com/nature/journal/v443/n7108/suppinfo/nature05113.html

References

   1. Enard, W. et al. Molecular evolution of FOXP2, a gene involved in speech and language. Nature 418, 869–872 (2002)

   2. Evans, P. D. et al. Adaptive evolution of ASPM, a major determinant of cerebral cortical size in humans. Hum. Mol. Genet. 13, 489–494 (2004)

   3. Hill, R. S. & Walsh, C. A. Molecular insights into human brain evolution. Nature 437, 64–67 (2005)

   4. Kimura, M. The Neutral Theory of Molecular Evolution (Cambridge Univ. Press, Cambridge, 1983)

   5. Chimpanzee Sequencing and Analysis Consortium. Initial sequence of the chimpanzee genome and comparison with the human genome. Nature 437, 69–87 (2005)

   6. Clark, A. G. et al. Inferring nonneutral evolution from human–chimp–mouse orthologous gene trios. Science 302, 1960–1963 (2003)

   7. Nielsen, R. et al. A scan for positively selected genes in the genomes of humans and chimpanzees. PLoS Biol. 3, e170 (2005)

   8. Dorus, S. et al. Accelerated evolution of nervous system genes in the origin of Homo sapiens. Cell 119, 1027–1040 (2004)

   9. Wang, X. , Grus, W. E. & Zhang, J. Gene losses during human origins. PLoS Biol. 4, e52 (2006)

  10. Waterston, R. H. et al. Initial sequencing and comparative analysis of the mouse genome. Nature 420, 520–562 (2002)

  11. King, M. C. & Wilson, A. C. Evolution at two levels in humans and chimpanzees. Science 188, 107–116 (1975)

  12. Pedersen, J. S. et al. Identification and classification of conserved RNA secondary structures in the human genome. PLOS Computat. Biol. 2, e33 (2006)

  13. Hillier, L. W. et al. Sequence and comparative analysis of the chicken genome provide unique perspectives on vertebrate evolution. Nature 432, 695–716 (2004)

  14. Lander, E. S. et al. Initial sequencing and analysis of the human genome. Nature 409, 860–921 (2001)

  15. Nobrega, M. A. , Ovcharenko, I. , Afzal, V. & Rubin, E. M. Scanning human gene deserts for long-range enhancers. Science 302, 413 (2003)

  16. Siepel, A. et al. Evolutionarily conserved elements in vertebrate, insect, worm, and yeast genomes. Genome Res. 15, 1034–1050 (2005)

  17. Hedges, S. B. & Kumar, S. Genomics. Vertebrate genomes compared. Science 297, 1283–1285 (2002)

  18. Altshuler, D. et al. A haplotype map of the human genome. Nature 437, 1299–1320 (2005)

  19. Collins, F. S. , Brooks, L. D. & Chakravarti, A. A. DNA polymorphism discovery resource for research on human genetic variation. Genome Res. 8, 1229–1231 (1998)

  20. Stajich, J. E. & Hahn, M. W. Disentangling the effects of demography and selection in human history. Mol. Biol. Evol. 22, 63–73 (2005)

  21. Merryman, C. & Noller, H. in RNA–Protein Interactions: A Practical Approach (ed. Smith, C.) 237–253 (Oxford Univ. Press, New York, 1998)

  22. Griffiths-Jones, S. et al. Rfam: Annotating non-coding RNAs in complete genomes. Nucleic Acids Res. 33, D121–D124 (2005)

  23. Meyer, G. & Goffinet, A. M. Prenatal development of reelin-immunoreactive neurons in the human neocortex. J. Comp. Neurol. 397, 29–40 (1998)

  24. Meyer, G. & Wahle, P. The paleocortical ventricle is the origin of reelin-expressing neurons in the marginal zone of the foetal human neocortex. Eur. J. Neurosci. 11, 3937–3944 (1999)

  25. Zecevic, N. & Rakic, P. Development of layer I neurons in the primate cerebral cortex. J. Neurosci. 21, 5607–5619 (2001)

  26. Meyer, G. , Soria, J. M. , Martinez-Galan, J. R. , Martin-Clemente, B. & Fairen, A. Different origins and developmental histories of transient neurons in the marginal zone of the fetal and neonatal rat cortex. J. Comp. Neurol. 397, 493–518 (1998)

  27. Rakic, S. & Zecevic, N. Emerging complexity of layer I in human cerebral cortex. Cereb. Cortex 13, 1072–1083 (2003)

  28. Perez-Garcia, C. G. , Tissir, F. , Goffinet, A. M. & Meyer, G. Reelin receptors in developing laminated brain structures of mouse and human. Eur. J. Neurosci. 20, 2827–2832 (2004)

  29. Lukaszewicz, A. et al. G1 phase regulation, area-specific cell cycle control, and cytoarchitectonics in the primate cortex. Neuron 47, 353–364 (2005)

  30. Bond, J. & Woods, C. G. Cytoskeletal genes regulating brain size. Curr. Opin. Cell Biol. 18, 95–101 (2006)

  31. Ptak, S. E. et al. Fine-scale recombination patterns differ between chimpanzees and humans. Nature Genet. 37, 429–434 (2005)

  32. Duret, L. , Semon, M. , Piganeau, G. , Mouchiroud, D. & Galtier, N. Vanishing GC-rich isochores in mammalian genomes. Genetics 162, 1837–1847 (2002)

  33. Bernardi, G. Isochores and the evolutionary genomics of vertebrates. Gene 241, 3–17 (2000)

  34. Kudla, G. , Lipinski, L. , Caffin, F. , Helwak, A. & Zylicz, M. High guanine and cytosine content increases mRNA levels in mammalian cells. PLoS Biol. 4, e180 (2006)

  35. Tissir, F. & Goffinet, A. M. Reelin and brain development. Nature Rev. Neurosci. 4, 496–505 (2003)

  36. Impagnatiello, F. et al. A decrease of reelin expression as a putative vulnerability factor in schizophrenia. Proc. Natl Acad. Sci. USA 95, 15718–15723 (1998)

  37. Alfano, G. et al. Natural antisense transcripts associated with genes involved in eye development. Hum. Mol. Genet. 14, 913–923 (2005)

  38. Lambot, M. A. , Depasse, F. , Noel, J. C. & Vanderhaeghen, P. Mapping labels in the human developing visual system and the evolution of binocular vision. J. Neurosci. 25, 7232–7237 (2005)

  39. Depaepe, V. et al. Ephrin signalling controls brain size by regulating apoptosis of neural progenitors. Nature 435, 1244–1250 (2005)

  40. de Bergeyck, V. , Naerhuyzen, B. , Goffinet, A. M. & Lambert de Rouvroit, C. A panel of monoclonal antibodies against reelin, the extracellular matrix protein defective in reeler mutant mice. J. Neurosci. Methods 82, 17–24 (1998)

  41. Kent, W. J. et al. The human genome browser at UCSC. Genome Res. 12, 996–1006 (2002)

What makes humans unique? A comparative genetics approach to this question reveals a candidate gene for a part in bestowing 'humanness'. Interestingly, the gene is expressed during embryonic development of the neocortex, the site of many of the brain's most sophisticated processes. The gene emerged from a scan for non-coding genomic regions that remained mostly unchanged up to the point of the chimp/human evolutionary split, but which then underwent accelerated evolution in humans. One of the most rapidly evolving genomic regions in the human lineage corresponds to a previously unstudied RNA gene, dubbed HAR1F. Gene expression studies then showed that HAR1F is active in cells called Cajal-Retzius neurons during the crucial period of gestation, when many of the nerve cells of the neocortex are establishing their functions within the brain.

Chris P. Ponting and Gerton Lunter

http://www.nature.com/nature/journal/v443/n7108/full/nature05154.html
doi:10.1038/nature05154

The differences in brain size and function that separate humans from other mammals must be reflected in our genomes. It seems that the non-coding 'dark matter' of genomes harbours most of these vital changes.

1. Lipovich L, Vanisri RR, Kong SL, Lin C-Y, and Liu ET,
"Primate-Specific Endogenous Cis-Antisense Transcription in the Human 5q31 Protocadherin Gene Cluster".

2. Zapala MA, Hovatta I, Ellison JA, Wodicka L, Del Rio JA, Tennant R, Tynan W, Broide RS, Helton R, Stoveken BS, Winrow C, Lockhart DJ, Reilly JF, Young WG, Bloom FE, Lockhart DJ, and Barlow C,
"Adult mouse brain gene expression patterns bear an embryologic imprint".

3. Kuwabara T, Hsieh J, Nakashima K, Taira K, and Gage FH, "A Small Modulatory dsRNA Specifies the Fate of Adult Neural Stem Cells". (2004) Cell, 116: 779-793.

Links to RNA and Biological Causality:

A Brief History of  Activator RNA:

Links to Euchromatin Activator RNA Reviews:
Links to Euchromatin Activator RNA Research:
Links to Ultrastructural Probes of DNase I-Sensitive Sites:
Links to RNA as a Therapeutic Agent:
Links to Hodgkin Lymphoma Immuno-Pathology:
Links to Activated T-Lymphocyte Immunotherapy:
Links to Medical Systems Biology:
Links to Selective Gene Transcription:
Links to RNA-Induced Epigenetics:
Links to RNA-Induced Embryogenesis:
Links to RNA and Biological Causality:
Links to Reprogramming and Neoplasia:

"Ultrastructural Probes of Active DNA Sites, and the RNA Activators of DNA".