Hemispheric asymmetry in new neurons in adulthood is associated with vocal learning and auditory memory

Abstract

Many brain regions exhibit lateral differences in structure and function, and also incorporate new neurons in adulthood, thought to function in learning and in the formation of new memories. However, the contribution of new neurons to hemispheric differences in processing is unknown. The present study combines cellular, behavioral, and physiological methods to address whether 1) new neuron incorporation differs between the brain hemispheres, and 2) the degree to which hemispheric lateralization of new neurons correlates with behavioral and physiological measures of learning and memory. The songbird provides a model system for assessing the contribution of new neurons to hemispheric specialization because songbird brain areas for vocal processing are functionally lateralized and receive a continuous influx of new neurons in adulthood. In adult male zebra finches, we quantified new neurons in the caudomedial nidopallium (NCM), a forebrain area involved in discrimination and memory for the complex vocalizations of individual conspecifics. We assessed song learning and recorded neural responses to song in NCM. We found significantly more new neurons labeled in left than in right NCM; moreover, the degree of asymmetry in new neuron numbers was correlated with the quality of song learning and strength of neuronal memory for recently heard songs. In birds with experimentally impaired song quality, the hemispheric difference in new neurons was diminished. These results suggest that new neurons may contribute to an allocation of function between the hemispheres that underlies the learning and processing of complex signals.

Figure 1. Experimental timeline.

Juvenile male zebra finches were exposed to adult male song throughout development (d = day from first BrdU injection). BrdU injections were administered 3x/day for 3 days. Four weeks after the last BrdU injection, multi-unit activity was recorded simultaneously from multiple electrodes placed bilaterally in NCM of awake, restrained birds in response to playback of auditory stimuli. Not all birds were used for all data sets: See text for sample sizes in each experimental condition.

Figure 2. New neurons in NCM.

(A) Medial (top, ∼170 µm from the midline) and lateral (bottom, ∼500 µm from the midline) sections of NCM (stippled area), showing the region where new neurons were quantified. Abbreviations: Cb, cerebellum; CMM, caudomedial mesopallium; HP, hippocampus; L2, Field L2; NCM, caudomedial nidopallium. Scale bar = 1 mm. (B) Photographs of a 1-month-old new neuron in NCM (arrow) in the same field of view showing a BrdU+ nucleus (top), Hu+ neuronal cytoplasm (middle) and co-localization of both markers (bottom). Scale bar = 10 µm.

Figure 3. Hemispheric asymmetry in new neurons.

(A) New neuron densities are significantly correlated between the hemispheres across birds. Points above the dashed identity line are birds with higher numbers of new neurons in left than right NCM. Best-fit regression (solid orange line) intercepts the Y-axis at 283.6 (orange arrow), indicating higher neuron density on the left (n = 31; r² = 0.644; P<0.001). Red symbols are computer-tutored birds. (B) 1-month-old new neuron densities in NCM differ between the hemispheres. Mean new neuron density/mm³ was higher in the left NCM than the right NCM (Left = 1340.85±179.04 SEM, Right = 1025.31±139.92 SEM; paired t-test; n = 31, t = 2.999, P = 0.005, two-tailed).

Figure 4. Hemispheric asymmetry in new neuron density is correlated with individual success in tutor song copying.

Higher new neuron density on the left than the right measured by a positive Neuron Asymmetry Index (NAI) is correlated with higher quality song imitation measured by the Similarity Index (n = 14; r² = 0.467; P = 0.007).

Figure 5. Vocal nerve section alters new neuron lateralization.

(A) New neuron densities in NCM differ between the hemispheres in a subset of control birds (Left = 1331±200 SEM, Right = 831±157 SEM; paired t-test; n = 7, t = 3.500, P = 0.013, two-tailed) but not in NXIIts-cut birds (Left = 793±149 SEM, Right = 933±174 SEM; paired t-test; n = 13, t = −0.545, P = 0.595 two-tailed). (B) The Neuron Asymmetry Index (NAI) shows a left-side bias in control birds (0.513±0.163 SEM, n = 7), significantly different from NAI in NXIIts-cut birds processed simultaneously (–0.277±0.253 SEM, n = 13; unpaired t-test, t = 2.133, P = 0.045).

Figure 6. Hemispheric difference in neuronal memory for songs heard in adulthood.

(A) Neuronal memory for songs heard 20 h earlier, measured as the Relative Response Strength (RRS) is significantly higher on the left than the right (n = 14; Left = 0.083±0.018 SEM; Right = 0.007±0.021 SEM; paired t-test; t = 3.95, P = 0.002, two-tailed). (B) Birds with the highest Neuron Asymmetry Index (NAI) indicating relatively more new neurons on the left side, had the highest RRS on the right. When responses in each hemisphere were compared to new neuron lateralization, NAI and RRS were not correlated in left NCM (n = 14; r² = 0.058; P = 0.406; white symbols). In contrast, RRS in right NCM had a wider range and was positively correlated with the NAI (n = 14; r² = 0.312, P = 0.038; black symbols).