Human TKTL1 implies greater neurogenesis in frontal neocortex of modern humans than Neanderthals

The evolutionary expansion of the neocortex and the concomitant increase in neuron production are considered to be a basis for the increase in cognitive abilities that occurred during human evolution. Endocast analyses reveal that the endocranial volume of modern humans and Neanderthals was similar, suggesting similar brain and neocortex size. But whether similar neocortex size implies similar neocortical neuron production remains unclear.

Transketolase-like 1 (TKTL1) is a gene from the transketolase family that in fetal human neocortex is preferentially expressed in the two classes of neuroprogenitors, the apical progenitors in the ventricular zone and the basal progenitors in the subventricular zone. The latter class of neuroprogenitors comprises two major types, the basal intermediate progenitors (bIPs) and the basal radial glia (bRG, also called outer radial glia). bRG exhibit cellular processes that promote their ability to self-amplify, and are the neuroprogenitor type considered to be a driver of the increase in cortical neuron production, which is a hallmark of the evolution of the human neocortex. Reflecting their cell polarity, bRG undergo repeated asymmetric divisions that self-renew the bRG and generate one neuron each. Thereby, bRG generate more neurons over time than the other type of neuron-generating basal neuroprogenitors, the process-lacking bIPs whose neurogenic divisions are symmetric self-consuming. TKTL1 is one of the few proteins with a single amino acid substitution found in essentially all present-day humans but absent from extinct archaic humans, the Neanderthals and Denisovans, and other primates. This human-specific amino acid substitution in TKTL1 is a lysine in apes and archaic humans but an arginine in modern humans. We therefore investigated (i) whether TKTL1 has a role in neocortex development and affects neuroprogenitor numbers and (ii) whether both archaic TKTL1 (aTKTL1) and modern human TKTL1 (hTKTL1) exert similar effects on neuroprogenitors during neocortex development.

When expressed in mouse embryo neocortex, which lacks TKTL1 expression, hTKTL1 increased the abundance of bRG without affecting that of bIPs and that of apical progenitors. The effect of TKTL1 on bRG abundance was limited to hTKTL1; aTKTL1, which differs only by one amino acid, was unable to increase bRG abundance. The greater bRG abundance upon hTKTL1 expression resulted in an increase in cortical neuron production over time, specifically of the late-born upper-layer neurons rather than of the early-born deep-layer neurons. In the folded (gyrencephalic) developing ferret neocortex, hTKTL1 expression increased not only bRG abundance but also the proportion of bRG with multiple processes, a hallmark of bRG that can self-amplify. As a consequence of this effect, gyrus size increased.

In fetal human neocortex, hTKTL1 was essential to maintain the full number of bRG, as CRISPR-Cas9–mediated hTKTL1 knockout in fetal human neocortical tissue reduced this number. To further demonstrate the relevance of this effect, we converted hTKTL1 to the Neanderthal variant aTKTL1 in human embryonic stem cells and generated minibrain structures called cerebral organoids. The aTKTL1-expressing organoids contained fewer bRG and neurons, hence the human-specific lysine-to-arginine substitution in hTKTL1 is essential for maintaining the full number of bRG and neurons in this human brain model. In fetal human neocortex, hTKTL1 expression in neuroprogenitors increased during the course of neurogenesis and was particularly high in the developing frontal lobe as compared to the developing occipital lobe.

As to its mechanism of action, hTKTL1 increased bRG abundance via two metabolic pathways, the pentose phosphate pathway (PPP) followed by fatty acid synthesis. Inhibition of the PPP or of fatty acid synthesis, using a variety of specific inhibitors, completely suppressed the hTKTL1-induced increase in bRG abundance in embryonic mouse neocortex and reduced bRG numbers in fetal human neocortical tissue. This metabolic action of hTKTL1, but not aTKTL1, in bRG resulted in an increase in the concentration of acetyl–coenzyme A, the critical metabolite for fatty acid synthesis. Our data suggest that hTKTL1, but not aTKTL1, promotes the synthesis of membrane lipids containing a certain type of fatty acid that are required for the outgrowth of bRG processes and hence for the increase in bRG abundance.

In light of our finding that TKTL1 expression in fetal human neocortex is particularly high in the developing frontal lobe, our study implies that because of the single amino acid–based activity of hTKTL1, neocortical neurogenesis in modern humans was and is greater than it was in Neanderthals, in particular in the frontal lobe.