The Journal of Neurophysiology Vol. 78 No. 5 November 1997, pp. 2817-2821
Copyright ©1997 by the American Physiological Society

RAPID COMMUNICATION

Gain Adaptation of Eye and Head Movement Components of Simian Gaze Shifts

James O. Phillips, Albert F. Fuchs, Leo Ling, Yoshiki Iwamoto, and Scott Votaw

Department of Physiology and Biophysics and Regional Primate Research Center, University of Washington, Seattle, Washington 98195

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Phillips, James O., Albert F. Fuchs, Leo Ling, Yoshiki Iwamoto, and Scott Votaw. Gain adaptation of eye and head movement components of simian gaze shifts. J. Neurophysiol. 78: 2817-2821, 1997. To investigate the site of gaze adaptation in primates, we reduced the gain of large head-restrained gaze shifts made to 50° target steps by jumping the target 40% backwards during a targeting saccade and then tested gain transfer to the eye- and head-movement components of head-unrestrained gaze shifts. After several hundred backstep trials, saccadic gain decreased by at least 10% in 8 of 13 experiments, which were then selected for further study. The minimum saccadic gain decrease in these eight experiments was 12.8% (mean = 18.4%). Head-unrestrained gaze shifts to ordinary 50° target steps experienced a gain reduction of at least 9.3% (mean = 14.9%), a mean gain transfer of 81%. Both the eye and head components of the gaze shift also decreased. However, average head movement gain decreased much more (22.1%) than eye movement gain (9.2%). Also, peak head velocity generally decreased significantly (20%), but peak eye velocity either increased or remained constant (average increase of 5.6%). However, the adapted peak eye and head velocities were appropriate for the adapted, smaller gaze amplitudes. Similar dissociations in eye and head metrics occurred when head-unrestrained gaze shifts were adapted directly (n = 2). These results indicated that head-restrained saccadic gain adaptation did not produce adaptation of eye movement alone. Nor did it produce a proportional gain change in both eye and head movement. Rather, normal eye and head amplitude and velocity relations for a given gaze amplitude were preserved. Such a result could be explained most easily if head-restrained adaptation were realized before the eye and head commands had been individualized. Therefore, gaze adaptation is most likely to occur upstream of the creation of separate eye and head movement commands.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Saccadic eye movements are very accurate even though they occur too rapidly to enlist visual feedback to guide the movement. Therefore, the various elements of the saccadic control system must be precisely calibrated to produce an accurate movement in the absence of such feedback. Since the number and efficacy of neurons and the properties of the peripheral motor apparatus may change throughout life, recalibration of these neural elements must be a continual process. Such a process would rely on visual information, in this case an evaluation of retinal error present at the end of the saccade. The characteristics of saccadic gain recalibration have been studied by using a paradigm introduced by McLaughlin (1967), who stepped a target backward as a saccade was made toward it. Several hundred repetitions of this paradigm gradually cause saccade amplitude to decrease so that the saccade often lands at the backstepped location (Fuchs et al. 1996; Straube et al. 1997). Saccades to ordinary target steps then fall short.

Several investigators have attempted to locate the neuronal site of these adaptive changes in the monkey. After visually guided saccades are adapted, other visually elicited saccades including delayed, memory-guided and express saccades also are adapted, suggesting that adaptation occurs downstream of the superior colliculus (Fuchs et al. 1996). This interpretation is consistent with single-unit (Goldberg et al. 1993) and stimulation studies (Melis and van Gisbergen 1996) in the colliculus of monkeys that had undergone saccadic gain reductions.

In all these studies, the monkeys' heads were restrained and the monkeys tracked targets only with their eyes. Under natural conditions, however, redirections of the line of sight (i.e., gaze shifts) are accomplished by combinations of eye and head movements. We wondered whether a gain reduction produced with head-restrained saccades might transfer to natural gaze shifts, where it might influence head and eye movements at several possible sites. If head-restrained adaptation caused gain reductions in the eye component alone, the adaptation locus would be at site 1 in the simple diagram of Fig. 1A. On the other hand, if adaptation caused gain reduction in both eye and head components, the adaptation locus could involve either both sites 1 and 2 or only site 3. We will now present data that adaptation does reduce both eye and head amplitude. Furthermore, a single gain adaptation at site 3 can account for all of our findings, while adaptation at sites 1 and 2 would have to involve a higher order process with matched but unequal gain changes at the two sites.

View larger version (23K): [in this window] [in a new window]   FIG. 1. A: schematic of the gaze control system and 3 possible sites where adaptation could occur. B: head-unrestrained gaze movements to 50° target steps before (· · ·) and after () head-restrained adaptation in experiment HM-15. Gaze (G), eye (E), and head (H) movements and their velocities (, , and ) are shown.

	METHODS

Abstract Introduction Methods Results Discussion References

Gaze, eye, and head movements were measured in three macaque monkeys. Eye position in space (Gaze, G) was measured with the search coil technique. A vertical post attached to the skull constrained head movement to the horizontal plane and served to engage a potentiometer for the measurement of head position (H). Monkeys were rewarded for directing their gaze at light-emitting diodes (LEDs) arranged on the horizon at a distance of 0.35 m. Recorded voltages proportional to gaze, head, and target position (LED location) were digitized off-line at 1 kHz. An analysis program calculated eye position in the head (E) as G-H, and identified the start and end of each movement component based on a 5°/s velocity criterion. The program also calculated the metrics, e.g., peak velocity, amplitude and duration of eye, head, and gaze movements. Our surgical techniques, monkey training procedures, target presentation, and methods of data reduction were published earlier (Phillips et al. 1995).

In 13 experiments, decreases in the amplitude of horizontal head-restrained saccades to 50° target steps were produced by jumping the target backward by 20° (by 40%) during a targeting saccade. Back-steps occurred in one target direction only, so that oppositely directed gaze shifts served as unadapted control movements. Target steps of 50° were initiated from pseudorandomly selected LED locations spaced at 5° but 30° eccentric to the midline of either the head (head-restrained condition) or the trunk (head-unrestrained condition), resulting in similar eye/gaze starting positions in all phases of each experiment. To optimize our chances for detecting the transfer of adaptation from head-restrained saccades to head-unrestrained gaze shifts we considered only the eight experiments in which the gain of the adapted head-restrained saccades had decreased by at least 10% after 1 h of adaptation trials. In two experiments we reduced gaze gain directly by jumping the target backward during the same number of 50° head-unrestrained gaze shifts.

Before and after adaptation, saccadic gain was assessed by requiring the monkeys to track an average of 35 ordinary target steps with no intrasaccadic back step. Similar numbers of head-unrestrained gaze shifts were compared before and after adaptation. To eliminate the gradual saccadic gain recovery that occurs in postadapted trials (Straube et al. 1997), we extinguished the target as the targeting saccade or gaze shift reached 40°/s in both the pre- and postadapted trials. Our quantitative analysis was performed only on data from such trials, and not on data collected during adaptation.

	RESULTS

Abstract Introduction Methods Results Discussion References

Head-restrained adaptation

Reductions of saccadic gain produced with head-restrained saccades showed considerable transfer to head-unrestrained gaze shifts. In the experiment illustrated in Fig. 1B, adaptation had reduced head-restrained saccades by 7.1° or by 15.7% (Table 1, HM-15). The average head-unrestrained gaze amplitude was reduced by 8.0° (15.5%), representing a gain transfer of 99% from head-restrained saccades. Across all eight experiments, the average transfer to gaze gainwas 81%.

Figure 1B also shows that the reduction of gaze gain was the result of reductions in both eye and head amplitude. Average eye amplitude decreased by 4.7° (12.2%), while average head amplitude decreased by 4.6° (19.1%; Table 1). The percentage reduction of head amplitude was greater than that of eye amplitude in four of five experiments in monkey HM and in all experiments on monkeys JB and TM (Table 1). Across all eight experiments, average head amplitude decreased by 5.2° (22.1%) while eye amplitude decreased by 3.3° (9.2%).

Head-restrained saccade adaptation also had differential effects on other metrics of eye and head movement. Peak eye velocity increased significantly (e.g., Fig. 2) in three of eight experiments after adaptation (Mann-Whitney U test, P  0.01), for an overall average increase in eight experiments of 18°/s (5.6%). In contrast, peak head velocity decreased significantly (P  0.01) in five of eight experiments, producing an overall average decrease of 24°/s (20%) for eight experiments. The duration of the eye movement decreased significantly after adaptation in seven of eight experiments (e.g., Fig. 1B) for an overall average of 29 ms in all eight experiments. In contrast, the duration of the head movement showed only a small average overall increase of 7 ms in the same experiments.

View larger version (36K): [in this window] [in a new window]   FIG. 2. Qualitative comparison of head-unrestrained and head-restrained gaze shifts before and after gain reduction by intrasaccadic target backsteps in experiment HM-10. Postadapted gaze shifts (B) matched to preadapted gaze shifts of either the same size (C) or to the same target amplitude (A). Thin lines: individual trials; thick lines: average movements.

Head-unrestrained adaptation

In two additional experiments in monkey JB, we adapted head-unrestrained gaze shifts directly by causing a backward target jump during the gaze shift. This paradigm produced an average 10.5% reduction in gaze amplitude, accompanied by average reductions in eye and head amplitude of 2.8 and 24.2%, respectively. Head-restrained saccades showed a 7.8% reduction in amplitude, reflecting a 75% transfer of adaptation. Therefore, adaptation of gaze movements with the head restrained or unrestrained produces disproportionate reductions of the eye and head components of a gaze shift.

Disproportionate reductions of eye and head components might seem unexpected if adaptation is occurring near the gaze command locus (site 3). However, adaptation near site 3 would produce proportional changes in eye and head components only if eye and head amplitude or eye and head velocity were linearly related for normal gaze shifts. In fact, for large gaze amplitudes, head amplitude increases when eye amplitude saturates and head velocity increases as eye velocity decreases (Freedman and Sparks 1997; Phillips et al. 1995) (Fig. 3, C and D). Therefore, for adaptation to occur at site 3, adapted eye and head movements need only have normal metrics. Indeed, when pre- and postadapted gaze shifts of similar amplitude (as in Fig. 2, B and C) are compared, they have similar time courses, as do their eye and head movement components. More complete data from another experiment illustrate that eye and head amplitude have the same relationship to gaze amplitude before (preadaptation) and after (postadaptation) adaptation (Fig. 3, A and B). Adaptation also did not alter the relationship between peak eye or head velocity and gaze amplitude (Fig. 3, C and D).

View larger version (32K): [in this window] [in a new window]   FIG. 3. Quantitative comparison of pre- and postadapted eye and head movements in experiment HM-7. A and B: eye and head amplitude vs. gaze amplitude. C and D: peak eye and head velocity vs. gaze amplitude. In this example, leftward (negative) gaze movements were not adapted.

To quantitatively assess these relationships, we compared amplitude matched (e.g., 35° ± 5°) pre- and postadaptation gaze shifts in all eight experiments. There were no significant differences in the pre- and postadaptation means of peak eye amplitude or velocity in any of the eight experiments, nor were there significant differences in the pre- and postadaptation means of peak head velocity or amplitude in seven of the eight experiments illustrated in Table 1 (Mann-Whitney U test, significant at P  0.01).

	DISCUSSION

Abstract Introduction Methods Results Discussion References

Our data indicate that gaze adaptation does not occur at site 1 alone since both eye and head movements are reduced. Furthermore, the gain reduction of head and eye movements is such that the adapted movements have the metrics expected of nonadapted gaze shifts of similar size. Such a reduction could easily be achieved by a simple parametric gain reduction at site 3. A more complicated combination of gain changes at sites 1 and 2 could produce similar results, but those changes would have to be precisely matched, and unequal. Therefore, we feel that the most parsimonious explanation for our data is that a simple gain adaptation occurred at site 3.

Most previous studies suggest that saccadic adaptation occurs near the motor output. These studies (e.g., Fuchs et al. 1996) looked for adaptation of different commands to the same motor system, the oculomotor system. In a sense, we took the reverse approach here, looking for adaptation across motor systems (eye and head) of a single command. When the question is posed in this manner, our results support the suggestion that adaptation influences a signal driving both the eye and the head, perhaps initial gaze error.

The interpretation of our data depends on the alternatives presented in the simple schematic in Fig. 1. While there is controversy as to the degree of independence of eye and head movement during a gaze shift, or the level in the nervous system at which the commands to eye and head diverge, there is little question that the initial input to a visually triggered gaze shift is gaze error (retinal error) or that the eye and head commands must be separate at some point. The scheme and alternatives presented in Fig. 1 are therefore rather conservative.

Finally, the observation that saccadic adaptation acts on both the eye and head components has profound implications for gaze control. If saccadic adaptation is the primary means of compensation for local disturbances in extra-ocular muscle force and/or changes in the gain of individual pathways, it appears to act through all of the participating motor systems. Thus the adaptive mechanisms do not simply restore a single faulty movement component, but rather restructure and restore the entire coordinated behavior.

	ACKNOWLEDGEMENTS

  This research was supported by National Institute of Health Grants RR-00166 and EY-00745.

	FOOTNOTES

  Address reprint requests to J. O. Phillips.

  Received 18 April 1997; accepted in final form 21 July 1997.

	REFERENCES

Abstract Introduction Methods Results Discussion References

• FREEDMAN, E. G., SPARKS, D. L. Eye-head coordination during head-unrestrained gaze shifts in rhesus monkeys. J. Neurophysiol. 77: 2328-2348, 1997.[Abstract/Free Full Text]
• FUCHS, A. F., REINER, D., PONG, M. Transfer of gain changes from targeting to other types of saccade in the monkey: constraints on the possible sites of saccadic gain adaptation. J. Neurophysiol. 76: 2522-2535, 1996.[Abstract/Free Full Text]
• GOLDBERG, M. E., MUSIL, S. Y., FITZGIBBON, E. J., SMITH, M., OLSON, C. R. The role of the cerebellum in the control of saccadic eye movements. In: Role of the Cerebellum and Basal Ganglia in Voluntary Movement, edited by N. Mano, I. Hamada, and M. R. Delong . Amsterdam: Elsevier, 1993, p. 203-211
• MCLAUGHLIN, S. C. Parametric adjustment in saccadic eye movements. Percept. Psychophys. 2: 359-362, 1967.
• MELIS, B.M.J., VAN GISBERGEN, J.A.M. Short-term adaptation of electrically induced saccades in monkey superior colliculus. J. Neurophysiol. 76: 1744-1758, 1996.[Abstract/Free Full Text]
• PHILLIPS, J. O., LING, L., FUCHS, A. F., SIEBOLD, C., PLORDE, J. J. Rapid horizontal gaze movement in the monkey. J. Neurophysiol. 73: 1632-1652, 1995.[Abstract/Free Full Text]
• STRAUBE, A., FUCHS, A. F., USHER, S., ROBINSON, F. R. Characteristics of saccadic gain adaptation in rhesus macaques. J. Neurophysiol. 77: 874-895, 1997.[Abstract/Free Full Text]

0022-3077/97 $5.00 Copyright ©1997 The American Physiological Society