Directional Bias of Neurons in V1 and V2 of Strabismic Monkeys: Temporal-to-Nasal Asymmetry?

Abstract

purpose. Strabismus that develops shortly after birth is known to cause temporal-to-nasal eye movement asymmetries under monocular viewing. The neural mechanisms underlying this deficit are not well understood. In the current study, the hypothesis that this eye movement anomaly reflects a similar asymmetry in the directional response properties of neurons in the early stages of cortical processing was examined.

methods. Strabismus was simulated with optical methods in infant monkeys between 4 and 14 weeks of age. When the monkeys were mature, microelectrode recording experiments were conducted in the primary visual cortex (V1) and visual area 2 (V2). After the spatial frequency of sine wave–grating stimuli for each neuron was optimized, each neuron’s responsiveness to 24 directions of stimulus movement was measured. The preferred direction and the strength of directional bias were determined by a vector summation method.

results. There was not an overabundance of neurons in V1 or V2 of strabismic monkeys preferring the temporal-to-nasal direction of stimulus movement. However, the average directional bias was significantly reduced in these strabismic monkeys. Interocular suppression was highly prevalent, and this suppression was stronger and more common in neurons dominated by the ipsilateral eye.

conclusions. The results suggest that the eye movement asymmetries in strabismic subjects do not result from similar asymmetries in the directional properties of V1 or V2 neurons, but rather reflect impoverished cortical signals to the brain stem nuclei that control eye movements.

Humans and monkeys with early-onset strabismus exhibit monocular nasotemporal asymmetries in pursuit eye movements ^{1 2} and visual tracking. ^{3 4} These asymmetries are closely associated with abnormal responses of neurons in the brain stem that control eye movements (e.g., the nucleus of optic tract [NOT]). For example, normally binocularly driven NOT neurons become monocular in cats and monkeys that experience early strabismus ^{5 6} or early monocular form deprivation. ^{7 8} Also, in binocularly form-deprived monkeys, there is a large decrease in the proportion of NOT neurons preferring the ipsiversive direction that is correlated with the measured temporal-to-nasal OKN asymmetry in these subjects. ⁸  

The abnormal physiology of NOT neurons and the resultant oculomotor asymmetries in strabismic subjects are thought to be a consequence of altered cortical input to the brain stem. ⁹ However, the nature of the altered signals from visual cortex to the brain stem is a matter of considerable debate. In studies of the early stages of cortical processing in which monocular (m)VEP was used, it was hypothesized that the oculomotor asymmetries reflect a similar nasotemporal asymmetry in visual motion processing by neurons in V1 and V2. ¹⁰ A competing hypothesis is that the nasotemporal asymmetry in smooth-pursuit eye movements in strabismic monkeys is caused by a weakness of ipsiversive input signals from MT/MST to the cortical pursuit system. ¹ These investigators found no asymmetry in the distribution of preferred direction among MT neurons. Moreover, human strabismic amblyopes exhibiting asymmetry in OKN had equal sensitivity to nasal and temporal stimulus motion, ¹¹ implying an absence of asymmetries in the motion sensitivity of neurons in the early stages of cortical processing. 

There are substantial projections from V1 and V2 down to the NOT complex in addition to the known connections from area MT/MST. ^{9 12} Also, mVEP signals recorded in the monkey brain largely reflect the neural activity of V1 and V2. ^{10 13} In this study, we used microelectrode recording methods to determine whether a temporal-to-nasal asymmetry in motion signal processing exists in V1 or V2 of strabismic monkeys. We also investigated whether the ocular dominance distribution is shifted in favor of the contralateral eye. To avoid adverse effects of potentially interfering with the normal maturation of the oculomotor system, we used optical methods to simulate early strabismus. 

Methods

All experimental procedures conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Subjects

The procedures for simulating strabismus by optical means have been described in detail elsewhere. ^{14 15 16} Briefly, four 4-week-old infant monkeys wore a pair of light-weight goggles that held 15-D prisms oriented base-in in front of each eye continually for a duration of 10 weeks, and the prism on the right eye was rotated base-down by 15° to ensure the absence of fusion during prism rearing. The total prismatic deviation exceeded the fusional vergence ranges of normal monkeys. The monkeys subsequently matured with unrestricted vision. 

At 18 months of age, spatial contrast-sensitivity functions were behaviorally measured in each eye in two of the four strabismic monkeys, by a standard operant method. ¹⁷ These two experimental monkeys showed normal spatial contrast-sensitivity functions in both eyes (i.e., no amblyopia), but there was no binocular summation of contrast sensitivity. No obvious ocular misalignment was detected in any of four strabismic monkeys. Refractive errors of all four experimental adult monkeys (+0.75, +1.0, +2.5, and +4.0 D) were within the range of refractive errors for normal adult monkeys measured around 3 years of age (i.e., between 0.0 and +6.0 D). ¹⁸ Although there was no measurement of ocular motility in our experimental monkeys, infant monkeys reared with similar, long-term prism-wear or naturally strabismic monkeys were reported to exhibit nasotemporal pursuit and OKN asymmetries. ¹⁹ In monkeys approximately 4 years of age, the microelectrode recording experiments were conducted in V1 and V2. Two monkeys served as normal adult control animals. 

Preparation

The surgical preparation and the recording and stimulation methods have been described in detail elsewhere. ^{20 21} Briefly, monkeys were anesthetized initially with an intramuscular injection of ketamine hydrochloride (15–20 mg/kg) and acepromazine maleate (0.15–0.2 mg/kg). A superficial vein was cannulated, and all subsequent surgical procedures were performed under sodium thiopental anesthesia. The animals were paralyzed by an IV infusion of pancuronium bromide (a loading dose of 0.1–0.2 mg/kg followed by a continuous infusion of 0.1–0.2 mg/kg per hour) and artificially respired with a mixture of 59% N₂O, 39% O₂, and 2% CO₂. Anesthesia was maintained by the continuous infusion of pentobarbital sodium (2–4 mg/k per hour). The core body temperature was kept at 37.6°C. Cycloplegia was produced by 1% atropine sulfate, and the animals’ corneas were protected with rigid, gas-permeable, extended-wear contact lenses. Retinoscopy was used to determine the contact lens parameters necessary to focus the eyes on the stimulus screen. 

Recording and Response Analysis

Tungsten-in-glass microelectrodes were used to isolate the activity from individual cortical neurons. For each isolated neuron, the receptive field for each eye was mapped with handheld stimuli. All receptive fields were located within 6.0° of the center of the fovea. The visual stimuli were generated in graphics software (Vision Research Graphics; Durham, NH) on a monochrome monitor (frame rate = 140 Hz; 800 × 600 pixels; mean luminance, 50 cd/m²). Responses to drifting sine wave gratings (3.1 Hz, 30%–40% contrast) were measured to determine the orientation and spatial frequency tuning functions for each unit (Figs. 1a 1b) . Cells were classified as simple or complex on the basis of the temporal characteristics of their responses to a drifting sine wave grating of the optimal spatial frequency and orientation. ²²  

Directional bias was calculated for each unit by a vector summation method. ²³ Specifically, stimuli of the optimal spatial frequency were drifted in 24 directions and each neuron’s response at each direction was represented by a vector, R _i = r exp(jθ _i ), where r is the scalar amplitude, and θ is the stimulus direction. Directional bias was then calculated by the following equation:  

\[\mathrm{Directional\ bias}\ {=}\ \frac{{\Sigma}R_{\mathrm{i}}/N}{{\Sigma}r_{\mathrm{i}}/N}\]

where ΣR _i is the vector sum for all 24 directions, Σr _i is the scalar sum for the amplitude of all R, and n = 24. 

The ocular dominance index (ODI) of each neuron was calculated using the following formula: ODI = (R _i − noise)/[(R _c − noise) + (R _i − noise)], where R _i is the peak response amplitude for ipsilateral eye stimulation, R _c is the peak response amplitude for contralateral eye stimulation, and noise is the spontaneous activity. Thus, an ODI of 0.0 represents a response exclusively favoring the contralateral eye, and an ODI of 1.0 represents a response exclusively favoring the ipsilateral eye. 

To determine the strength and the nature of binocular interactions, responses were collected for dichoptic sine wave gratings of the optimal spatial frequency and orientation as a function of the relative interocular spatial phase disparity of the grating pair (Fig. 1c) . ^{21 24} The sensitivity to relative interocular spatial phase disparities was quantified with a binocular interaction index: BII = amplitude of the fitted sine wave/the average binocular response amplitude. To characterize whether binocular signal interactions were facilitatory or suppressive in nature, the peak binocular response amplitude/dominant monocular response amplitude ratios (peak B/M ratios) were calculated for each unit and expressed in terms of relative strength (decibels) that is, 10 log peak B/M. Negative peak B/M ratios signify binocular suppression, and positive B/M ratios indicate binocular facilitation. 

At the end of each penetration, small electrolytic lesions (5 μA, 5 seconds, electrode negative) were made at three points along the track both in V1 and V2 for later reconstruction). Experiments were terminated by administering an overdose of pentobarbital sodium (100 mg/kg), and the animals were killed by perfusion through the heart with an aldehyde fixative. Frozen sections were stained for Nissl substance and cytochrome oxidase. 

Results

The quantitative analysis of directional bias was performed in two normal control monkeys (n = 53 for V1 and n = 253 for V2) and four strabismic monkeys (n = 144 for V1 and n = 199 for V2). Typical penetrations were made through the operculum of V1, approximately 3 mm posterior to the lunate sulcus and 15 to 20 mm lateral from the midline. The angle of each penetration was kept nearly constant for all monkeys, so that the electrode traversed all the layers of V1 for a distance of approximately 2.5 to 3.0 mm, followed by a traverse of the white matter for approximately 600 μm, after which the electrode entered layer 6 of V2 and traversed the remaining 5 layers, for a distance of approximately 4 to 5 mm. 

Uniform Distribution of Preferred Stimulus Direction

There was no obvious asymmetry in the distribution of preferred directions in either V1 or V2 of our strabismic monkeys. Each data point in Fig. 2shows the preferred direction in angular coordinates and the strength of a unit’s preference (i.e., directional bias). The polar plots for V1 and V2 in both strabismic and normal monkeys show a relatively uniform distribution of directional preferences. There was no obvious overrepresentation of any preferred direction (Moore’s test, P > 0.05). 

More important, there was no systematic temporal-to-nasal asymmetry in theses polar plots (Figs. 2 3) . When the mean directional bias of those units falling within 22.5° of the eight primary directions was compared, there was no significant preference for any direction (ANOVA, P > 0.1). Moreover, when the mean directional bias of those units with preferred stimulus directions within ±45° of either nasal or temporal direction (Fig. 3 ; arrowheads) was calculated, there was no significant asymmetry in V1 or in V2 (t-test, P > 0.1). 

Reduced Directional Bias in Strabismic Monkeys

Although we did not find any evidence of nasotemporal asymmetry in directional preferences of individual neurons, the overall directional bias of V1 and V2 neurons was clearly reduced in strabismic monkeys (t-test, P < 0.001; Fig. 4 ). This decrease in directional bias was found for all directions, when the mean directional biases for those units falling within 22.5° of the eight primary directions were compared (ANOVA, P < 0.05; Fig. 3 ). Moreover, if the mean directional bias of those units exhibiting preferred stimulus directions within ±45° of either nasal or temporal directions were analyzed (Fig. 3 , arrowheads), there was a significant reduction in the mean directional bias of both V1 and V2 neurons in strabismic monkeys compared to normal control animals (t-test, P < 0.01). These results suggest that V1 and V2 neurons of strabismic monkeys exhibit a significant loss in their ability to signal stimulus motion in all directions, including both temporal and nasal directions. 

Cortical projections to the NOT originate primarily from the infragranular layers of visual cortical areas. ^{9 25} Therefore, it is possible that neurons in these layers of strabismic monkeys may exhibit a temporal-to-nasal asymmetry in directional bias. ^{13 26} Polar plots in Figure 5 , illustrating the preferred direction and the directional bias of each unit in the infragranular layers (layers 5 and 6), show that there was no asymmetry that favored the nasalward stimulus motion. (Moore’s test, P > 0.05.) If the mean directional bias of those units exhibiting preferred stimulus directions within ±45° of either nasal or temporal directions were analyzed (Fig. 5 , open arrows), there was no asymmetry in the mean directional bias for either V1 or V2 neurons (t-test, P > 0.1). 

However, it is important to note that the overall directional bias of infragranular layer neurons in V1 and V2 was significantly reduced in strabismic monkeys (t-tests, P < 0.01), and that the mean directional bias of those units exhibiting the preferred stimulus direction within ±45° of either nasal or temporal directions was also significantly lower in strabismic monkeys than in normal control animals (Fig. 5 , white arrows, t-tests, P < 0.05 in V1, P < 0.001 in V2). These results suggest that the ability of these brain-stem–projecting neurons of V1 and V2 to signal stimulus movement in either nasal or temporal directions were significantly reduced in strabismic monkeys. 

Binocular Response Properties and Interocular Suppression

The ocular dominance distribution of V1 and V2 neurons in Figure 6demonstrates important changes in the binocularity of neurons in strabismic monkeys. Specifically, we found a predictable loss of binocularly driven units in V1 and a small shift toward the contralateral eye. To our surprise, we did not observe a comparable reduction in the percentage of binocularly driven V2 neurons. Instead, a relatively large proportion of V2 units responded well to stimulation of either eye. This is in sharp contrast to a severe loss of binocularly driven MT neurons in monkeys reared with early onset, surgically induced strabismus and a dramatic ocular dominance shift in favor of the contralateral eye. ¹ However, note that the ocular dominance distribution of V2 neurons in our strabismic monkeys also exhibited a modest but clear shift toward the contralateral eye (χ² test, P = 0.001). 

Binocular interaction experiments with dichoptic pairs of gratings yielded several important observations. Although the ocular dominance tests indicated that V2 neurons in strabismic monkeys received excitatory drive from either eye, signals from the two eyes were not combined in a manner that supported disparity sensitivity. The mean BII values of both V1 and V2 neurons in our strabismic monkeys were significantly lower than those in normal control monkeys (t-test, P < 0.001; Fig. 7a ). 

A condition in which cortical drive to the NOT would be weakened in strabismic monkeys would be if cortical neurons showed anomalous binocular suppression during early development and did not drive NOT neurons as strongly as in normal infants. A consequence would be that the developing functional connections from visual cortex to the NOT would be weakened and cortical control over the NOT would also be diminished. We found that the mean peak binocular-over-monocular response ratio (peak B/M) was significantly reduced in strabismic monkeys (t-tests, P < 0.001), and that the percentage of units exhibiting binocular suppression (peak B/M ratios < 0.0 dB) was significantly increased (Mann-Whitney test, P < 0.01; Fig. 7b ), indicating the presence of robust binocular suppression in V1 and V2. 

More revealing results of binocular suppression are illustrated in Figure 8 , in which the average peak B/M ratios (Fig. 8a)and the proportion of binocularly suppressive units (Fig. 8b)are shown as a function of the ocular dominance indices. In V2 of strabismic monkeys, there was a significant increase in the proportion of units showing interocular suppression (χ² test, P < 0.001) and a significant decrease in the average peak B/M ratio (t-test, P < 0.001) in binocularly balanced (ODI = 0.4–0.6) and ipsilaterally dominating units (ODI > 0.6). The extremely low encounter rate of ipsilaterally dominated V2 units (ODI > 0.8) in strabismic monkeys compared to that in V1 (Fig. 5)may be related to the strong suppression of signals originating in the contralateral eye. 

Discussion

The main finding of this study was that neither V1 nor V2 neurons in monkeys reared with early-onset strabismus showed a temporal-to-nasal asymmetry in signaling directions of stimulus movement. Instead, we found that V1 and V2 neurons in strabismic monkeys were generally not as direction-selective as those in normal monkeys, and that during binocular viewing, interocular suppression was highly prevalent in both V1 and V2 neurons. More important, this suppression was stronger and more common in neurons dominated by the ipsilateral eye. 

Lack of Nasotemporal Asymmetry

The primary evidence of nasotemporal asymmetry in cortical motion processing of neonates and strabismic subjects comes from a series of mVEP studies. ^{10 13 19 26 27 28 29} In these studies VEP responses to contrast-reversing gratings were dominated by the first harmonic component, and the responses were 180° out of phase in the two eyes during monocular viewing. These abnormal mVEP responses in normal infants and strabismic subjects were hypothesized to reflect a high prevalence of units exhibiting nasal or temporal direction preference and/or variations in the strength of direction tuning for cells selective for nasal versus temporal motion in the early stages of cortical processing. The present results and our previous findings in neonates ³⁰ are inconsistent with either of these possibilities. The present results in both V1 and V2 of strabismic monkeys are similar to the previous observation that the preferred directions of MT neurons in monkeys reared with convergent strabismus are uniformly distributed. ¹  

The source of the apparent disagreement between these studies is unclear. One possibility is that a nasotemporal bias may be present just for those units that project to the NOT. ²⁶ The distribution of directional bias for infragranular neurons in strabismic monkeys (Fig. 5)did not support this hypothesis. Also, it is highly unlikely that a small percentage of neurons in the deep cortical layers in V1 and V2 that project to the brain stem would give rise to strong bias in the mVEP, especially if the most of the units in the more superficial layers of the same cortical areas do not show the temporal-to-nasal bias. 

Although there was no nasotemporal asymmetry in the distribution of preferred direction, the quantitatively determined directional biases of V1 and V2 units were significantly reduced in strabismic monkeys. More important, neurons in the infragranular laminas also showed similar reductions in the average directional bias. These results suggest that the overall influence of cortical signals over the directional bias of NOT neurons is likely to have been altered in strabismic monkeys because, in addition to the known projections from area MT, ^{1 31} the major cortical inputs to the NOT originate from V1 and V2. Unresolved questions are how much of the directional bias of NOT neurons is influenced by cortical signals relative to direct retinal inputs and which cortical area has the greatest influences on the responses of NOT neurons. 

Ocular Dominance Shift and Binocular Suppression

In a previous study, early strabismus in infant monkeys was shown to cause a severe asymmetry in smooth-pursuit eye movements and a loss of binocularly driven units in the MT accompanied by a large-scale shift in the ocular dominance distribution toward the contralateral eye. ¹ There was no evidence of a temporal-to-nasal asymmetry in the pure motion-processing of area MT and these investigators concluded that the observed eye movement asymmetry was due to a weakness in the ipsiversive input from area MT to the cortical pursuit system caused by the ocular dominance shift. 

In addition to the lack of asymmetry in the overall directional bias of V1 and V2 neurons, our results on the binocular responses of V1 and V2 neurons are generally consistent with the findings in area MT. However, the ocular dominance test in this study revealed a larger proportion of binocularly balanced V2 neurons than units highly dominated by one eye (Fig. 5) . We also found a milder shift in ocular dominance distribution toward the contralateral eye. 

More important, we found a high prevalence of interocular suppression that was stronger in ipsilaterally dominated units (Fig. 8) . This result is consistent with the earlier anatomic observations that, although early strabismus does not alter the widths of ocular dominance columns of macaque V1, ³² cytochrome oxidase activity was reduced in ocular dominance columns driven by the ipsilateral eye of monkeys that had infantile esotropia. ³³ These alterations may be closely associated with the overall reduction in ipsiversive influences over the brain stem nuclei by V1 and V2 neurons. 

Conclusions

The oculomotor asymmetries in strabismic monkeys are not a direct consequence of a nasotemporal asymmetry in visual motion processing by an assembly of V1 and V2 units. Instead, a loss of direction selectivity and a high prevalence of binocular suppression during early development weaken the ipsilateral functional connections from visual cortex to brain stem nuclei. Hence, the responses of the NOT neurons become monocular and are dominated by direct retinal input that favors the temporal-to-nasal motion over that in the opposite direction. To support this idea, however, the functional maturation of the NOT neurons must be investigated in normal infant monkeys, to determine how and at what age cortical activity begins to influence the normal development of the receptive field properties of neurons in the brain stem. 

 Supported by National Eye Institute Grants EY08128, EY03611, EY07751, and RR-07146.

 Submitted for publication May 9, 2005; revised June 24, 2005; accepted August 23, 2005.

 Disclosure: I. Watanabe, None; H. Bi, None; B. Zhang, None; E. Sakai, None; T. Mori, None; R.S. Harwerth, None; E.L. Smith III, None; Y.M. Chino, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Example tuning functions of a V2 neuron from a normal adult monkey. (a) Direction–orientation tuning, (b) spatial frequency tuning, and (c) binocular phase tuning. R bias and L bias, directional bias for the right and left eyes, respectively; Dom, dominant monocular response; Non-Dom, nondominant monocular response; Noise, spontaneous discharge; BII, binocular interaction index; peak B/M, peak monocular response amplitude/monocular response amplitude. Dotted line: mean binocular response.

Figure 2.

 

Absence of nasotemporal asymmetries in V1 and V2 of strabismic monkeys. Scatterplots illustrating the directional preference and the directional bias of individual V1 and V2 neurons for normal and strabismic monkeys. Note the decrease in the proportion of V1 and V2 neurons that exhibited a high directional bias in strabismic monkeys. N, nasal direction; T, temporal direction.

Figure 3.

 

Absence of temporal-to-nasal asymmetries and reductions in overall directional bias of V1 and V2 neurons in strabismic monkeys compared with that in normal monkeys. Mean ± SE directional biases of those units falling within 22.5° of the eight primary directions. Arrowheads: mean directional biases of those units exhibiting the preferred stimulus direction within ±45° of either nasal or temporal directions. N, nasal direction; T, temporal direction.

Figure 4.

 

Reduction in directional bias of V1 and V2 neurons in strabismic monkeys. Mean ± SE directional bias of all units for V1 and V2 neurons of normal and strabismic monkeys.

Figure 5.

 

Lack of nasotemporal asymmetry in directional bias of the brain-stem–projecting neurons in strabismic monkeys. Scatterplots illustrate the directional preference and the directional bias of individual V1 and V2 neurons in infragranular layers (layers 5 and 6) of normal and strabismic monkeys. Unfilled arrows: mean directional biases of those units that exhibited the preferred stimulus direction within ±45° of either the nasal or temporal direction.

Figure 6.

 

Increase in the proportion of neurons preferring the contralateral eye stimulation in strabismic monkeys. ODI distribution of all V1 and V2 neurons in normal and strabismic monkeys. An ODI of 0.0 represents a response exclusively favoring the contralateral eye, and an ODI of 1.0 represents a response exclusively favoring the ipsilateral eye.

Figure 7.

 

Reductions of disparity sensitivity and increases of interocular suppression in V1 and V2 of strabismic monkeys. (a) Mean ± SE sensitivity of V1 and V2 neurons to interocular spatial phase disparity in normal and strabismic monkeys. (b) Comparisons of the peak binocular-over-monocular response ratio of units in V1 and V2 in normal and strabismic monkeys. Each horizontal bar indicates a data point for a single unit. (○) Median values; (□) mean values. The proportion of suppressive units (i.e., peak B/M < 0.0 dB) in each group is shown at the bottom.

Figure 8.

 

Binocular suppression as a function of a neuron’s ODI, demonstrating increased suppression in V2 neurons of strabismic monkeys that were dominated by the ipsilateral eye. Mean ± SE peak binocular over monocular response ratios (in decibels) as a function of ODI for V1 (a) and V2 (b) neurons in normal and strabismic monkeys. Proportion of binocularly suppressive V1 (c) and V2 (d) units as a function of ODI in normal and strabismic monkeys. ( ) The number of units was too small (n = 4) to make meaningful comparisons for this group.