Multimodal analysis in infants to adults shows faster myelination of V1 than high-level visual areas

2141 

Abstract Submission 

Clara Bacmeister¹, Vaidehi Natu², Mercedes Paredes^3,4, Kalanit Grill-Spector²

¹Stanford University, Stanford, CA, ²Psychology Department, Stanford University, Stanford, CA, ³Department of Neurology, University of California - San Francisco, San Francisco, CA, ⁴Eli and Edythe Broad Center of Regeneration Medicine and Stem Cell Research, University of California, San Francisco, CA

Clara Bacmeister  
Stanford University
Stanford, CA

Vaidehi Natu  
Psychology Department, Stanford University
Stanford, CA

Mercedes Paredes  
Department of Neurology, University of California - San Francisco|Eli and Edythe Broad Center of Regeneration Medicine and Stem Cell Research, University of California
San Francisco, CA|San Francisco, CA

Kalanit Grill-Spector  
Psychology Department, Stanford University
Stanford, CA

Myelination is a lifelong process that can be modified by experience and exert profound effects on circuit function, learning, and memory. Yet, how myelination contributes to the development and maturation of human visual circuits is unknown.

Here, we examined the development of myelin using a combination of ex vivo histology (immunohistochemistry, IHC) (Fig. 1a) and in vivo quantitative MRI (qMRI) of tissue relaxation rate (R₁) in three functionally-distinct regions of human visual cortex: primary visual cortex (V1) and high-level face- and place-selective cortex. We chose these regions as they include both early and high-level visual areas that can be identified using anatomical landmarks alone in ex-vivo samples (Holmes 1918, Weiner 2014; Weiner 2018). IHC of myelin basic protein (MBP) was done in 6 infant samples (34 gestational weeks to 5 months old), a 9 year old, and a 25 year old. Analysis of myelin coverage was conducted using semi-automated thresholding of MBP+ immunolabeling to calculate total cortical area covered by MBP+ myelin sheaths. Cross-sectional and longitudinal qMRI was done in 80 infants (newborn-12 month olds) and cross-sectional qMRI was done in 52 5-25 year olds. In adults, R₁ in cortex depends on myelin (Gomez 2017; Natu 2018) and the physiochemical tissue properties (Mezer 2013; Edwards 2018).

Using IHC, we find that myelin increases in infancy in all regions, but more rapidly in the calcarine sulcus (V1), than in the collateral sulcus (CoS; place-selective) and fusiform gyrus (FG; face-selective) (Fig. 1b-d). Furthermore, while myelin levels in V1 and CoS are similar in childhood and adulthood (9 vs. 25 years old), in the FG, myelin levels are lower in the 9 year old than in the adult, suggesting continued myelination of FG during adolescence (Fig. 1b). Within visual areas analyzed, cortical layers display different developmental trajectories of myelination, where deep layers of cortex (L4-L6) and L1 myelinate earlier and more rapidly than L2/3. In the case of higher-level visual cortex (CoS and FG), both L1 and L2 continue to myelinate throughout adolescence. Cortical R₁ is higher in V1 than higher-level face- and place- selective cortex at birth, and all regions show faster increases in R₁ in infancy than later childhood (Fig. 2a), mirroring myelin patterns in ex vivo samples. Across the lifespan, cortical R₁ plateaus earliest in V1. In contrast, cortical R₁ reaches adult levels in place-selective cortex in childhood (5-9 years old), and cortical R₁ is still lower in 10-12 year old children than in adults in face-selective cortex, suggesting prolonged development of face-selective cortex into adulthood (Fig 2b). Across ages and areas, there is a positive linear relationship between cortical R₁ and myelin level estimated from IHC (beta=0.0069,p<0.0001,no effect of region or interaction), suggesting that in cortex linear increases in myelin will produce linear increases in R₁. As we find visual cortex is largely devoid of myelin at birth, we identify a baseline cortical R₁ value (0.49 s^-1) that is not driven by myelin.

Overall, we observe differential development of cortical myelin, where myelination proceeds rapidly during the first year of life and continues developing more slowly during childhood, with differential development across layers and areas. Deeper layers myelinate earlier than superficial layers, and V1 myelinates earlier than high-level regions, with face-selective cortex showing the most prolonged myelination. We confirm that developmental changes in R₁ are correlated to changes in myelin and identify baseline cortical R₁ value that is independent of myelin. These findings identify differences in spatiotemporal patterning of myelination within the human visual system, laying the foundation to understanding differences in functional development.

Normal Brain Development: Fetus to Adolescence

Cortical Cyto- and Myeloarchitecture ¹

Multi-Modal Imaging

Perception: Visual ²

Cortex

Cortical Layers

Development

MRI

Myelin

NORMAL HUMAN

Vision

White Matter