Transient visual pathway critical for normaldevelopment of primate grasping behaviorInaki-Carril Mundinanoa, Dylan M. Foxa, William C. Kwana, Diego Vidaurreb, Leon Teoa, Jihane Homman-Ludiyea,Melvyn A. Goodalec, David A. Leopoldd, and James A. Bournea,1

aAustralian Regenerative Medicine Institute, Monash University, Clayton, VIC 3800, Australia; bOxford Centre for Human Brain Activity, Department ofPsychiatry, University of Oxford, Oxford OX3 7JX, United Kingdom; cThe Brain and Mind Institute, The University of Western Ontario, London, ON, CanadaN6A 5B7; and dLaboratory of Neuropsychology, National Institute of Mental Health, Bethesda, MD 20892

Edited by Peter L. Strick, University of Pittsburgh, Pittsburgh, PA, and approved December 4, 2017 (received for review September 27, 2017)

An evolutionary hallmark of anthropoid primates, includinghumans, is the use of vision to guide precise manual movements.These behaviors are reliant on a specialized visual input to theposterior parietal cortex. Here, we show that normal primatereaching-and-grasping behavior depends critically on a visualpathway through the thalamic pulvinar, which is thought to relayinformation to the middle temporal (MT) area during early life andthen swiftly withdraws. Small MRI-guided lesions to a subdivisionof the inferior pulvinar subnucleus (PIm) in the infant marmosetmonkey led to permanent deficits in reaching-and-grasping be-havior in the adult. This functional loss coincided with theabnormal anatomical development of multiple cortical areas re-sponsible for the guidance of actions. Our study reveals that thetransient retino–pulvinar–MT pathway underpins the develop-ment of visually guided manual behaviors in primates that arecrucial for interacting with complex features in the environment.

prehension | pulvinar | marmoset | thalamus | visual cortex

The expansion of the visual cortex in primates has been ac-companied by the elaboration of intracortical projection

systems, manifest as partially segregated dorsal and ventral visualstreams (1). Hierarchies of areas within the two streams arethought to mediate the accurate and rapid transformation ofvisual information into complex percepts and behaviors. Thedorsal visual stream is specialized to mediate visually guidedbehaviors such as reaching and grasping. It draws largely uponinput from middle temporal (MT) area, which relays dynamicvisual signals to posterior parietal cortex, where areas are knownto coordinate visual input with primate-specific actions (2, 3)(Fig. 1A). The dorsal visual stream is established early in life, andits areal components mature rapidly compared with other cor-tical areas, including the primary visual cortex. This early de-velopment is thought to support certain behaviors crucial forneonatal survival, with early experience in turn shaping thefunctional capacities of the parietal cortex (4–8). Recently, themechanisms underlying this early dorsal stream maturation havebeen linked to a transient visual pathway that is present at birthand that involves direct retinal innervation of the pulvinar. Inthis pathway, the retinorecipient medial subdivision of the in-ferior pulvinar (PIm) relays visual signals to cortical area MTduring early life (4, 9) (Fig. 1B). Then over the course of severalmonths, the geniculostriate visual pathway strengthens its con-nections to MT as the retino–pulvinar pathway diminishes.In the present study, we set out to measure the causal effects

of this transient developmental pathway on dorsal-streamstructure and function. Specifically, we hypothesized that earlydisruption of the retino–pulvino–MT pathway removes a criticalsource of visual input for the experience-dependent maturationof the parietal areas involved in the visual guidance of manualmovements (10, 11) (Fig. 1C). We further hypothesized thatthese changes would be associated with measurable alterations invisually guided manual behaviors. To investigate these hypoth-eses, we performed precise, early-life unilateral lesions of PIm in

marmoset monkeys, thus disrupting the retino–pulvinar pathwayto MT in one hemisphere. We subsequently analyzed the con-sequences of these lesions in the adult, examining alterations inreach-and-grasp behaviors, longitudinal changes in cortical dif-fusion MRI, and biochemical markers of cortical maturation.

ResultsEstablishment of PIm Lesions. Using a precision MRI-guided ap-proach (12), we performed unilateral chemical lesions of PIm at14–22 d after birth (Fig. 1C), a period when the anatomicalretino–pulvinar–MT pathway is most prominent, before its sub-sequent pruning (4). Lesions were restricted to the left hemisphere,which in primates has been specifically linked to visuomotor control(13, 14). We used three independent methods to verify the efficacyof the lesion, which were successfully localized to PIm. First, MRIT2-weighted images showed darker hypointense signals in the PImlesion sites (Fig. 1D). Second, darkened MRI regions were colo-calized with a reduction in the expression of the neuronal markerNeuN, compared with the unlesioned PIm in the contralateralhemisphere, as well as other pulvinar subdivisions (Fig. 1 E andF and Fig. S1). Third, retrograde tracing of the PIm–MT pathwayrevealed almost complete degeneration of the pathway (Fig. 1G and H). Together, these results indicated a successful excitotoxiclesion of cells that were largely confined to PIm.

Abnormal Reach-and-Grasp Behavior. To assess the long-termfunctional consequences of early PIm lesions, we measuredand quantified aspects of visually guided grasping in the adult(>18 mo), comparing those having undergone early-life PIm

Significance

Visually guided actions have been central to the neural and be-havioral evolution of primates, culminating in precision graspingand object manipulation. Here, we reveal that the establishmentand maturation of the cortical areas controlling reach-and-graspbehaviors are dependent on a transient visual pathway from theretina to the medial inferior pulvinar of the thalamus, whichrelays the information to the dorsal stream of the visual cortex,via the middle temporal (MT) area, in early life. Disruption of thispathway in early life leads to permanent anatomical, connec-tional, and reach-and-grasp behavior deficits.

Author contributions: I.-C.M., D.M.F., M.A.G., D.A.L., and J.A.B. designed research; I.-C.M.,D.M.F., W.C.K., and J.A.B. performed research; L.T., J.H.-L., and J.A.B. contributed newreagents/analytic tools; I.-C.M., D.M.F., W.C.K., D.V., L.T., J.H.-L., M.A.G., D.A.L., and J.A.B.analyzed data; and I.-C.M., D.V., L.T., M.A.G., D.A.L., and J.A.B. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Published under the PNAS license.

See Commentary on page 1143.1To whom correspondence should be addressed. Email: james.bourne@monash.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1717016115/-/DCSupplemental.

1364–1369 | PNAS | February 6, 2018 | vol. 115 | no. 6 www.pnas.org/cgi/doi/10.1073/pnas.1717016115

Dow

nloa

ded

by g

uest

on

Dec

embe

r 17

, 201

9

lesions (n = 2) to intact control animals (n = 2). During theexecution of a reach-to-grasp movement, the posture of the handis gradually adjusted to reflect the contours of the target object.As this behavior, known as preshaping, has not previously beenstudied in the marmoset, we developed a naturalistic protocoland applied a video-based kinematic analysis of the hand andarm movements to test whether dynamic parameters of visuallyguided reaching and grasping in the adult are influenced by theearly retino–pulvinar–MT pathway (Fig. 2A, Fig. S2A, and MovieS1). Our behavioral analysis focused on measuring the maximumgrip aperture (MGA) as a measure of hand preshaping, velocityand acceleration as a measure of reaching, as well as the complexdynamics of visually guided reaching within the context of thedual-component model. This model divides reaching into twocomponents, an initial feedforward phase related to planningprocesses and a second feedback phase related to processes thatmediate hand path corrections (15).The results revealed that the unilateral PIm-lesioned animals

had significant bimanual deficits compared with the controlgroup. General performance, as well as individually analyzed

reach-and-grasp components of the behavior, were significantlydifferent between the two groups (permutation testing, value ofP < 0.0001) (Fig. S2 B–D). Specifically, PIm-lesioned animalshad a significantly larger MGA in the preshaping phase (Fig. 2A and B, Fig. S2 D and E, and Movie S1) and approachedtargets with faster reach movements than controls (Fig. 2C andFig. S2C).Consistent with the crossed nature of the descending cortical

motor connections (16), certain aspects of the lesion-induceddeficit were asymmetric, and primarily affected the contrale-sional (right) hand, which was the preferred hand for all animals(see Table S1 and SI Materials and Methods for hand preferenceassessment). For example, the right-arm reaches in the PIm-lesioned animals exhibited aberrant temporal dynamics; thePIm-lesioned animals showed a protracted period of decelera-tion, consistent with a disruption to normal visually mediatedfeedback phase (Fig. 2D). The capacity to learn such visuallyguided movements with the dominant hand also was affected bythe early lesion. Whereas control animals readily improved theirperformance on the reach-to-grasp task over multiple repetitions,

Fig. 1. Transient medial inferior pulvinar (PIm) connectivity and the outcome of an early-life unilateral lesion. (A) Diagram showing the connectivity of themarmoset retinorecipient PIm with cortical areas controlling reach-and-grasping behaviors. Retino–pulvino–MT pathway areas, white in-fill; other directlyconnected visual cortical areas, gray in-fill. Note that the retinogeniculostriate projection is not indicated. (B) Illustration of the transient retino–PIm–MTpathway. Between postnatal day 7 (P7) and P30, this pathway is prominent but regresses after P90 when the direct V1–MT connection becomes moreprominent (4). (C) Hypothesis of the study. Neurotoxic ablation of PIm in early life would disrupt the normal maturation of dorsal-stream areas and associatedbehaviors. (D) T2-weighted MRI revealing the lesion site in the left PIm. Inset shows higher-magnification image of hatched area. Lesion outlined in yellow.(E) NeuN expression was clearly reduced in the lesioned PIm. Extent of lesion is demarcated with yellow hatched line. (Scale bar, 500 μm.) (F) High-magni-fication photomicrographs showing reduced NeuN expression in the lesioned PIm compared with the intact (Middle) PIm and medial pulvinar (PM; Right).(Scale bar, 50 μm.) (G) Neuronal tracing images confirming the ablation of the PIm–MT pathway (n = 2). PIm is demarcated by white hatch line. (Scale bar,200 μm.) (H) Box-and-whisker plot showing the reduced density of MT projecting neurons in lesioned PIm (Mann–Whitney U test, P = 0.0286, n = 4).

Mundinano et al. PNAS | February 6, 2018 | vol. 115 | no. 6 | 1365

NEU

ROSC

IENCE

SEECO

MMEN

TARY

Dow

nloa

ded

by g

uest

on

Dec

embe

r 17

, 201

9

the lesioned animals exhibited virtually no learning (Fig. S2F).Together, these deficits in the PIm-lesioned animals indicateabnormal visually mediated cortical control of reaching-and-grasping behavior.

Longitudinal Diffusion Tensor Imaging Analysis.Given the profoundreaching-and-grasping deficits, we wanted to examine the struc-tural changes in the visual and association cortex during theperiod after lesion into adulthood that might underlie the ob-served phenotype. Diffusion tensor imaging (DTI) is a form ofMRI that detects the Brownian motion of water molecules in avoxel and has the ability to determine microstructural abnor-malities and neuropathology in the cortex (17). Therefore, fol-lowing the PIm lesion, we tracked the progression of hemisphericstructural asymmetry, comparing lesioned (left) and unlesioned(right) hemispheres (n = 4), in the lesioned animals at 6, 9, 18,and 37 wk after the lesion. Using a voxel-based morphometricanalysis of DTI maps, we found that early PIm lesions led to aprogressive reduction in fractional anisotropy (FA) across sev-eral dorsal visual-stream areas ipsilateral to the lesion at 18 wkpostlesion (p.l.) (Fig. 3A). These FA changes were most prom-inent in areas of the association cortex related to reach-and-grasp visuomotor control, such as anterior intraparietal (AIP),medial intraparietal (MIP), and posterior parietal (PE) cortices(18–20).Longitudinal region of interest (ROI) analysis of FA, axial

diffusivity (AD), and radial diffusivity (RD) maps identified thatthe changes in FA and were evident by 6 wk p.l. and reachedstatistical significance (Mann–Whitney U test) only at 9 wk p.l.

(n = 4, P = 0.021), 18 wk p.l. (n = 4, P = 0.021), and 37 wk p.l.(n = 4, P = 0.021) (Fig. 3 B–D). A similar significance was ob-served in the AD maps (Mann–Whitney U test) only at 9 wk p.l.(n = 4, P = 0.021) and 18 wk p.l. (n = 4, P = 0.043) (Fig. S3 A andB), while RD maps remained unaltered at all stages (Mann–Whitney U test, P = 0.2 at 6 and 37 wk p.l.; P = 0.343 at 9 and18 wk p.l.).The marmoset cortical tissue is rich in structures organized

perpendicular to the surface of the cortex, such as axonal bun-dles and apical dendrites of pyramidal neurons (e.g., Fig. 4). Thiswas reflected in our DTIs, as tensors had a principal axis that wasalso perpendicular to the cortical surface. The reduced FA indorsal-stream areas ipsilateral to the lesion correlated with adecreased AD, which corresponds to the diffusivity along themain longitudinal axis of the tensor (Fig. S3C). Hence, FA andAD changes might indicate microstructural alterations in thenormal development of cortical elements with a perpendicularorientation to the cortical surface.

Decreased Anatomical Projections from Area MT to Parietal Cortex.In light of the noticeable changes to dorsal-stream areas ob-served by DTI, we next used conventional tract-tracing injectionsto ask whether this apparent disruption in parietal cortex de-velopment was reflected in the density of anatomical projections.In agreement with previous descriptions, our tracing injectionsshowed that area MT efferents directly innervated parietal as-sociation areas AIP, MIP, and PE (21, 22) (Fig. 4). Consistentwith the hypothesis of abnormal maturation of the dorsal path-ways, histological analysis in adulthood revealed that the terminal

Fig. 2. Early-life PIm ablation results in persistentdeficits in reach to grasp task performance. (A) Rep-resentative images taken from kinematical analysis ofthe reach-and-grasp task. Red line between thumband index finger denotes grip aperture. Note largermaximum grip aperture (MGA) in lesioned animal(arrow). (B) Probability distribution histogram acrosstrials of grip aperture for control and lesioned animals.Lesioned animals exhibit increased MGA (permutationtesting, P ≤ 0.0001). (C) Probability distribution show-ing increased velocity and acceleration of reach in le-sioned animals compared with controls (permutationtesting, P ≤ 0.0001). (D) Detailed analysis of reach dy-namics. The histograms show probability distributionfor control and lesion animals for both hands. Theprofile for the contralesional hand of PIm lesion ani-mals revealed a protracted period of deceleration(permutation testing, value of P < 0.0001).

1366 | www.pnas.org/cgi/doi/10.1073/pnas.1717016115 Mundinano et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 17

, 201

9

fields originating from MT in these areas were markedly less densein the lesioned animals (n = 2) compared with control animals (n =2) (Fig. 4 and Fig. S4). This observation suggests that the transientvisual pathway through PIm to MT is instrumental in the estab-lishment of MT efferent connections with these dorsal-stream areasin early life.

Disrupted Neuronal Maturation. Finally, we exploited known bio-chemical markers of neuronal maturation to investigate how, specif-ically, the parietal areas were affected by early PIm lesions. We foundthat, in the lesioned animals, the pyramidal neuron maturationmarker, nonphosphorylated neurofilament (NNF), was expressed byfewer cells in ipsilesional area MT (layer 3) and areas AIP, MIP, andPE (layer 5) (Fig. 5) (Mann–Whitney U test, n = 4, P = 0.021 for alldensity comparisons). This decrease in the NNF expression existed,despite a normal overall density of neurons in each of these areas(Fig. S5A). Furthermore, the remaining NNF-immunopositive neu-rons in these areas of the lesioned hemisphere were significantlysmaller in terms of their soma size compared with those in theunlesioned hemisphere (Fig. S5B). This observation correlates withthe microstructural alterations revealed by the decreased FAand AD values described above and is indicative of abnormalcircuit development (5, 23). We similarly quantified the par-valbumin (PV) interneuron population in the same areas be-cause of their known role in modulating the excitatory responseof pyramidal neurons (24), and their association with sensoryexperience and maturation of the cortex (25). We found thatthe PV-immunopositive neuron density was reduced in the same

areas and layers as NNF (Fig. 5) (Mann–Whitney U test, n = 4,P = 0.021 for all density comparisons), suggestive of alterations in thelocal circuit and potential for dysfunction (26).

Fig. 3. Diffusion tractography imaging (DTI)changes in dorsal-stream areas postlesion. (A) voxel-based morphometry of fractional anisotropy (FA)maps. Yellow-red clusters represent significanthemispheric differences in FA, which were lower inthe lesioned hemisphere and predominantly in theparietal cortex (Left vs. Right, n = 4). Vertical lines (1–4) on Inset 3D brain (Bottom Right corner) representinteraural level of the sections on Left side of thepanel. (B) Representative FA maps of a PIm-lesionedmarmoset 18 wk p.l. at level of the dorsal streamvisual cortex. FA values lower than 0.2 within ROIsare shown in yellow, while FA values higher than0.2 are shown in blue. (C) Box-and-whisker plotsshowing FA changes between left and right hemi-sphere dorsal-stream ROIs at different time points.(D) Graph showing the longitudinal progression ofthe FA changes.

Fig. 4. Disruption of PIm–MT relay in early life alters connectivity of dorsal-stream areas into adulthood. Heat maps revealing reduced connectivity(terminal fields) between MT and reach-and-grasp–related areas of thedorsal stream in the lesioned hemisphere. (Scale bar, 200 μm.)

Mundinano et al. PNAS | February 6, 2018 | vol. 115 | no. 6 | 1367

NEU

ROSC

IENCE

SEECO

MMEN

TARY

Dow

nloa

ded

by g

uest

on

Dec

embe

r 17

, 201

9

DiscussionTogether, these results provide clear evidence that eliminatingthe transient retino–pulvinar input to MT in early life rapidlydisrupts development of multiple dorsal-stream areas, resultingin the dislocation of the typical network integration and leading

to a sustained defect in visually guided actions. This transience isa common motif in embryonic and postnatal development, insome cases affecting cell populations (e.g., cortical subplate)(27), and in others affecting projection pathways (e.g., layer5 projection neurons) (28).The altered behaviors observed in this study are akin to those

observed in human patients with optic ataxia. This neurologicalsyndrome appears as a consequence of lesions to the posteriorparietal cortex and results in deficits in visually guided reachingand grasping, even though visual perception remains reasonablyintact (29). Furthermore, optic ataxia patients often present withpoor preshaping when reaching to grasp (30), opening their handmuch wider than normal like that described here for PIm-lesioned marmosets. Similar to what we observed followingthe unilateral lesion of PIm in the marmoset, bimanual deficitsafter unilateral left cortical lesions are also common in humanapraxia patients (31). Our results, therefore, suggest that the early

Fig. 5. Altered cellular maturation of dorsal-stream areas associated withreach-and-grasp behaviors. The pyramidal neuron maturation marker NNFwas significantly down-regulated in the lesioned hemisphere of areas MT(layer 3), AIP, PE, and MIP (layer 5) compared with the unlesioned hemi-sphere. The interneuronal marker PV was also down-regulated in the sameareas and cortical layers. (Scale bar, 100 μm.) White hatched lines indicatelayer boundaries. Insets show higher-magnification images of pyramidalneurons. (Scale bar, 50 μm.) Box-and-whisker plots (Right column) showingsignificant reduction in NNF and PV neuronal density for dorsal-stream areasfollowing a PIm lesion.

Fig. 6. Evolution of the retino–PIm–MT pathway in the anthropoid pri-mates. Our results suggest that retinal input to the PIm is essential for thedevelopment of the dorsal-stream visual cortex and posterior parietal cortexareas associated with reach-and-grasp behavior. The visual input to thepulvinar nucleus is conserved in mammals, but it is in the common ancestorof anthropoids when a very specialized retinal input to PIm appears for thefirst time. As a consequence, a much more complex area MT+ evolved, en-abling the expansion of the dorsal-stream visual cortex and posterior pari-etal cortex (PPC). This inflection point in primate evolution might beinvolved in the development of precise reach-and-grasp behaviors in theanthropoid primates, culminating in the precision reach-and-grasp behaviorsobserved in humans.

1368 | www.pnas.org/cgi/doi/10.1073/pnas.1717016115 Mundinano et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 17

, 201

9

retino–pulvinar–MTpathway is specific for the initial shaping of visualguidance of hand action and in the establishment of the dorsal-streamnetwork that is independent of the primary geniculostriate pathway.The implication of these findings extends our understanding of

primate evolution. One of the key adaptations of primates is thevisually guided action of the hand, which has been linked toreaching and grasping branches, insects, and terminal fruitswithin an arboreal environment (32, 33). The pulvinar nucleushas evolved considerably in the primates, with the direct retinalinput to PIm first emerging in the immediate ancestor of anthropoidmonkeys (34), being absent in the prosimians. This event coincideswith the expansion of area MT, and its complex connectivity withthe dorsal-stream visual areas, including the posterior parietal re-gions (35) (Fig. 6). The emergence of an early and transient retinal–pulvinar–MT pathway in the simians, facilitating use-dependentmaturation of the dorsal stream and its visual guidance of thehand, is likely to be a unique and an important adaptation thatfundamentally shaped primate behavior and cognition.

Materials and MethodsA total of eight New World marmoset (Callithrix jacchus) monkeys wereemployed in this study. All experiments were conducted in accordance with

the Australian Code of Practice for the Care and Use of Animals for ScientificPurposes and were approved by the Monash University Animal EthicsCommittee, which also monitored the welfare of these animals. Four neo-natal animals (14–22 d postnatal) received a chemical lesion to the lefthemisphere medial subdivision of the PIm. Animals were allowed to recoveruntil adulthood (18 mo) before we conducted neural tracing and post-mortem histological and immunohistochemical studies (n = 2; both female),trained and tested them on behavioral tasks, and carried out postmortemhistological and immunohistochemical studies (n = 2; one female, one male).All PIm-lesioned animals were included in MRI/DTI longitudinal analysisperformed at 6, 9, 18, and 37 wk p.l.. A further four intact adult marmosets(18–22 mo) were used as controls for the neural tracing (n = 2; one female,one male) and behavioral (n = 2; both females) studies. Materials andmethods are described in detail in SI Materials and Methods.

ACKNOWLEDGMENTS. We thank Qi Zhu Wu for his assistance with MRIacquisition and DTI preprocessing, and personnel of Monash BiomedicalImaging, Toby Merson, Greg Stuart, and Peter Currie for helpful advice onearlier drafts. The Australian Regenerative Medicine Institute is supportedby grants from the State Government of Victoria and the AustralianGovernment. This work was supported by National Health and MedicalResearch Council (NHMRC) Project Grant APP1042893. D.A.L. is supported bythe Intramural Research Program of the National Institute of Mental Healthunder Grants ZIA MH002838 and ZIA MH002898. J.A.B. is supported byNHMRC Senior Research Fellowship APP1077677.

1. Goodale MA, Milner AD (1992) Separate visual pathways for perception and action.Trends Neurosci 15:20–25.

2. Kaas JH, Gharbawie OA, Stepniewska I (2011) The organization and evolution ofdorsal stream multisensory motor pathways in primates. Front Neuroanat 5:34.

3. Kaas JH, Stepniewska I (2016) Evolution of posterior parietal cortex and parietal-frontal networks for specific actions in primates. J Comp Neurol 524:595–608.

4. Warner CE, Kwan WC, Bourne JA (2012) The early maturation of visual cortical areaMT is dependent on input from the retinorecipient medial portion of the inferiorpulvinar. J Neurosci 32:17073–17085.

5. Mundinano IC, Kwan WC, Bourne JA (2015) Mapping the mosaic sequence of primatevisual cortical development. Front Neuroanat 9:132.

6. Gogtay N, et al. (2004) Dynamic mapping of human cortical development duringchildhood through early adulthood. Proc Natl Acad Sci USA 101:8174–8179.

7. Kiorpes L, Price T, Hall-Haro C, Movshon JA (2012) Development of sensitivity toglobal form and motion in macaque monkeys (Macaca nemestrina). Vision Res 63:34–42.

8. Wattam-Bell J, et al. (2010) Reorganization of global form and motion processingduring human visual development. Curr Biol 20:411–415.

9. Warner CE, Goldshmit Y, Bourne JA (2010) Retinal afferents synapse with relay cellstargeting the middle temporal area in the pulvinar and lateral geniculate nuclei.Front Neuroanat 4:8.

10. Krubitzer LA, Kaas JH (1990) Cortical connections of MT in four species of primates:Areal, modular, and retinotopic patterns. Vis Neurosci 5:165–204.

11. Palmer SM, Rosa MG (2006) Quantitative analysis of the corticocortical projections tothe middle temporal area in the marmoset monkey: Evolutionary and functionalimplications. Cereb Cortex 16:1361–1375.

12. Mundinano I-C, Flecknell PA, Bourne JA (2016) MRI-guided stereotaxic brain surgeryin the infant and adult common marmoset. Nat Protoc 11:1299–1308.

13. Gonzalez CLR, Ganel T, Goodale MA (2006) Hemispheric specialization for the visualcontrol of action is independent of handedness. J Neurophysiol 95:3496–3501.

14. Serrien DJ, Ivry RB, Swinnen SP (2006) Dynamics of hemispheric specialization andintegration in the context of motor control. Nat Rev Neurosci 7:160–166.

15. Desmurget M, Grafton S (2000) Forward modeling allows feedback control for fastreaching movements. Trends Cogn Sci 4:423–431.

16. Lemon RN, Kirkwood PA, Maier MA, Nakajima K, Nathan P (2004) Direct and indirectpathways for corticospinal control of upper limb motoneurons in the primate. ProgBrain Res 143:263–279.

17. Le Bihan D, et al. (2001) Diffusion tensor imaging: Concepts and applications. J MagnReson Imaging 13:534–546.

18. Gharbawie OA, Stepniewska I, Kaas JH (2011) Cortical connections of functional zonesin posterior parietal cortex and frontal cortex motor regions in new world monkeys.Cereb Cortex 21:1981–2002.

19. Stepniewska I, Gharbawie OA, Burish MJ, Kaas JH (2014) Effects of muscimol inacti-

vations of functional domains in motor, premotor, and posterior parietal cortex on

complex movements evoked by electrical stimulation. J Neurophysiol 111:1100–1119.20. Murata A, Gallese V, Luppino G, Kaseda M, Sakata H (2000) Selectivity for the shape,

size, and orientation of objects for grasping in neurons of monkey parietal area AIP.

J Neurophysiol 83:2580–2601.21. Majka P, et al. (2016) Towards a comprehensive atlas of cortical connections in a

primate brain: Mapping tracer injection studies of the common marmoset into a

reference digital template. J Comp Neurol 524:2161–2181.22. Cerkevich CM, Collins CE, Kaas JH (2014) Cortical inputs to the middle temporal visual

area in new world owl monkeys. Eye Brain 2015:1–15.23. Bourne JA, Rosa MG (2006) Hierarchical development of the primate visual cortex, as

revealed by neurofilament immunoreactivity: Early maturation of the middle tem-

poral area (MT). Cereb Cortex 16:405–414.24. Atallah BV, Bruns W, Carandini M, Scanziani M (2012) Parvalbumin-expressing in-

terneurons linearly transform cortical responses to visual stimuli. Neuron 73:159–170.25. Chattopadhyaya B, et al. (2004) Experience and activity-dependent maturation of

perisomatic GABAergic innervation in primary visual cortex during a postnatal critical

period. J Neurosci 24:9598–9611.26. Le Magueresse C, Monyer H (2013) GABAergic interneurons shape the functional

maturation of the cortex. Neuron 77:388–405.27. Bystron I, Blakemore C, Rakic P (2008) Development of the human cerebral cortex:

Boulder committee revisited. Nat Rev Neurosci 9:110–122.28. Stanfield BB, O’Leary DDM, Fricks C (1982) Selective collateral elimination in early

postnatal development restricts cortical distribution of rat pyramidal tract neurones.

Nature 298:371–373.29. Andersen RA, Andersen KN, Hwang EJ, Hauschild M (2014) Optic ataxia: From Balint’s

syndrome to the parietal reach region. Neuron 81:967–983.30. Jakobson LS, Archibald YM, Carey DP, Goodale MA (1991) A kinematic analysis of

reaching and grasping movements in a patient recovering from optic ataxia.

Neuropsychologia 29:803–809.31. Kimura D, Archibald Y (1974) Motor functions of the left hemisphere. Brain 97:

337–350.32. Cartmill M (1992) New views on primate origins. Evol Anthropol 1:105–111.33. Sussman RW, Tab Rasmussen D, Raven PH (2013) Rethinking primate origins again.

Am J Primatol 75:95–106.34. Baldwin MKL, Balaram P, Kaas JH (2017) The evolution and functions of nuclei of the

visual pulvinar in primates. J Comp Neurol 525:3207–3226.35. Baldwin M, Bourne JA (2016) The evolution of subcortical pathways to the extra-

estriate cortex. Evolution of Nervous Systems, ed Kaas JH (Academic, Oxford).

Mundinano et al. PNAS | February 6, 2018 | vol. 115 | no. 6 | 1369

NEU

ROSC

IENCE

SEECO

MMEN

TARY

Dow

nloa

ded

by g

uest

on

Dec

embe

r 17

, 201

9