This study examined the question of which features of a complex grouping discrimination make it vulnerable to permanent elimination by V4 lesions. We first verified that the line element grouping discrimination, which we previously reported to be devastated by V4 lesions, was similarly affected in the monkeys of this study. The permanence of the deficit was established by mapping its visual field distribution and then testing this discrimination for an extended period at a locus on the border of the deficit. Also, a staircase procedure was used to provide the monkey with within session instruction in the grouping discrimination, but this did not improve V4 lesion performance. Grouping was then compared with several discriminations that shared some features with it, but which were found not to be permanently eliminated by V4 lesions. This comparison suggested that grouping (rather than segmentation or response to a single element) was one feature that made the discrimination vulnerable, a second was the similarity in shape of the texture elements to be grouped. Finally, we tested visual crowding, a property of peripheral vision that is thought to reflect neuronal interactions early in visual cortex, possibly in area V1, and found no effect of V4 lesions. A control experiment with human observers tested whether the elimination of grouping by V4 lesions might be due to an alteration of attention, but found no evidence to support this hypothesis. These results show that severe disruption of texture discriminations by V4 lesions depends on both the nature of the discrimination and the type of texture elements involved, but does not necessarily involve the disruption of attention.

Pursuit and perception both require accurate information about the motion of objects. Recovering the motion of objects by integrating the motion of their components is a difficult visual task. Successful integration produces coherent global object motion, while a failure to integrate leaves the incoherent local motions of the components unlinked. We compared the ability of perception and pursuit to perform motion integration by measuring direction judgments and the concomitant eye-movement responses to line-figure parallelograms moving behind stationary rectangular apertures. The apertures were constructed such that only the line segments corresponding to the parallelogram's sides were visible; thus, recovering global motion required the integration of the local segment motion. We investigated several potential motion-integration rules by using stimuli with different object, vector-average, and line-segment terminator-motion directions. We used an oculometric decision rule to directly compare direction discrimination for pursuit and perception. For visible apertures, the percept was a coherent object, and both the pursuit and perceptual performance were close to the object-motion prediction. For invisible apertures, the percept was incoherently moving segments, and both the pursuit and perceptual performance were close to the terminator-motion prediction. Furthermore, both psychometric and oculometric direction thresholds were much higher for invisible apertures than for visible apertures. We constructed a model in which both perception and pursuit are driven by a shared motion-processing stage, with perception having an additional input from an independent static-processing stage. Model simulations were consistent with our perceptual and oculomotor data. Based on these results, we propose the use of pursuit as an objective and continuous measure of perceptual coherence. Our results support the view that pursuit and perception share a common motion-integration stage, perhaps within areas MT or MST.

This issue of Visual Neuroscience contains the last scientific contribution of Brian B. Boycott. In December 1999, Brian was just recovering from a serious operation in London, when I sent him an early draft of the paper. He returned the manuscript with many excellent suggestions for improvements and in the accompanying letter he wrote, “I think you are generous to include my name. I won't argue because when it comes out in 2000 it will be the 50th anniversary of my first published paper. This appeals to me.” Unfortunately Brian did not live to see the publication of this paper and passed away on April 22, 2000.

It has been previously demonstrated that the majority of the glutamatergic input to directionally selective (DS) ganglion cells in the rabbit retina is mediated by NMDA receptors. To examine whether NMDA channels have any role in directional selectivity, we eliminated magnesium from the superfusion medium to prevent the magnesium block of the channels at hyperpolarized membrane potentials. During superfusion in magnesium-free media, the response to null-direction motion increased to the level of the response to preferred-direction motion. This effect was specifically mediated by NMDA channels because subsequent blocking of the NMDA channels with AP7 restored directional selectivity. We also tested whether the increase in the null-direction response in magnesium-free medium was due to an increased release of acetylcholine from the cholinergic amacrine cells, rather than an effect on the DS ganglion cells themselves, by blocking acetylcholine transmission with d-tubocurarine during superfusion with the magnesium-free medium. During zero-magnesium superfusion, d-tubocurarine reduced both the preferred- and null-direction responses of DS ganglion cells but did not restore directional selectivity. These findings suggest that null-direction motion normally causes portions of the dendritic membrane of the directionally selective ganglion cell to be maintained at a sufficiently negative potential that the NMDA channels are blocked by magnesium ions. This result is discussed in terms of several models for the mechanisms of directional selectivity.

In five subjects, we measured visual evoked potentials (VEPs) elicited by Vernier targets in which the contrast of the two components of the stimuli were modulated by sinusoids at distinct frequencies f1 and f2. This approach allows for the extraction of VEP signatures of spatial interactions, namely, responses at intermodulation frequencies n1f1 + n2f2, without the need to introduce motion into the stimulus. The most prominent interactions were at the sum frequency f1 + f2, and, for frequency pairs that were sufficiently separated, the difference frequency f1 − f2. These responses had a systematic dependence on the temporal parameters of the stimulus, corresponding to an effective latency of 145 to 165 ms. Fourth-order interactions were also detected, particularly at the frequencies 2f1 ± 2f2. These VEP signatures of interaction were similar to interactions seen for colinear line segments separated by a gap. Thus, for Vernier stimuli devoid of motion, VEP signatures of interaction are readily detected but are not specific to hyperacuity displacements. The distribution of interactions across harmonic orders is consistent with local rectification preceding the spatial interactions. Their effective latencies and dependence on spatial parameters are consistent with interactions within V1 receptive fields or mediated by horizontal connections between cells with a similar orientation tuning within V1.

(article appeared in Visual Neuroscience (1999), 16, 1105–1114.)

Due to an oversight, Table 2 that appeared on p. 1106 of Visual Neuroscience 16:6 had errors in the body of the Table. A corrected version of Table 2 is printed below.

Cambridge University Press and the authors regret any inconvenience that this inadvertent error may have caused.

The retinal degeneration B (RdgB) protein family is characterized by an amino-terminal phosphatidylinositol transfer protein (PITP) domain, several hydrophobic domains, and a highly conserved carboxyl terminus. We identified a zebrafish RdgB homolog (pl-RdgB) that lacks the amino-terminal PITP domain, while retaining over 45% amino acid identity with the two mouse RdgB proteins (M-RdgB1 and M-RdgB2). Unlike the widespread retinal expression observed for other vertebrate RdgB homologs, pl-RdgB is restricted in the retina to the cone cell inner segments. The pl-RdgB protein is also expressed in the brain, although its distribution is different than the other RdgB homologs. Analogous to M-RdgB2, pl-RdgB protein is extracted from a retinal homogenate by guanidine and not by Triton X-100. Thus, pl-RdgB and likely all the identified RdgB homologs are not integral membrane proteins, but may associate with the membrane through protein–protein interactions. While expression of either murine RdgB homolog restored the defective light response and prevented retinal degeneration in rdgB mutant flies, expressing zebrafish pl-RdgB in Drosophila rdgB2 null mutants slowed retinal degeneration without restoring the electrophysiological light response. Thus, pl-RdgB may define a previously unrecognized protein family, which includes the other RdgB homologs, that act through a protein complex to maintain photoreceptor viability.

We have examined the patterns of projections from different nuclei of the brain stem to the ventral lateral geniculate nucleus (vLGN) of the thalamus. Injections of biotinylated dextran were made into different nuclei of the brainstem (i.e., midbrain reticular nucleus, pontine reticular nucleus, deep layers of superior colliculus, periaqueductal grey matter [ventrolateral, dorsolateral, and lateral columns], pedunculopontine tegmental nucleus, parabrachial nucleus, lateral dorsal tegmental nucleus, substantia nigra [pars reticulata], locus coeruleus, and dorsal raphe) of Sprague-Dawley rats using stereotaxic coordinates. Our results show that all of the abovementioned brain-stem nuclei have overlapping projections to the medial regions of the vLGN, within the parvocellular lamina of the nucleus. This if the first instance of the parvocellular lamina being shown to receive a major set of projections. Very few labelled terminals from the brain stem were ever seen within the larger more lateral magnocellular lamina, which has been shown by previous studies to receive heavy inputs from visually associated structures, such as the retina and occipital cortex. Since many of the brain-stem nuclei injected in this study have little to do with visual processing, our results suggest that one can perhaps package the vLGN into distinct visual (magnocellular) and nonvisual (parvocellular) components.

(article appeared in Visual Neuroscience (1999), 16, 1145–1156.)

Due to an unfortunate misunderstanding, Figure 3 that appeared on p. 1149 of Visual Neuroscience 16:6 was not sized to the full two-column measure. A corrected version of the article, with Figure 3 reshot to full size, is reproduced on pp. 1145R–1156R, which follows. Cambridge University Press regrets any inconvenience that this inadvertent error may have caused.