Effects of Betaxolol on Light Responses and Membrane Conductance in Retinal Ganglion Cells

Abstract

purpose. To examine the physiological effects of betaxolol, aβ ₁-adrenergic receptor blocker commonly used in the treatment of glaucoma, on retinal ganglion cells and to evaluate its potential to elicit responses consistent with a neuroprotective agent against ganglion cell degeneration.

methods. Single-unit extracellular recording, electroretinogram (ERG), intracellular and whole-cell patch-clamp recording techniques were made from flatmounted, isolated retina, superfused eyecup, and living retinal slice preparations of the larval tiger salamander.

results. Bath application of 20 μM betaxolol reduced the glutamate-induced increase of spontaneous spike rate in retinal ganglion cell by approximately 30%. The glutamate-induced postsynaptic current recorded under voltage-clamp conditions was reduced by 50 μM betaxolol, and the difference current-voltage (I-V) relation (I_Control -I_betaxolol) was N-shaped and AP5-sensitive, characteristic of N-methyl-d-aspartate receptor–mediated current. Application of 50 μM betaxolol reversibly reduced the voltage-gated sodium and calcium currents by approximately one third of their peak amplitudes. The times-to-action of betaxolol on ganglion cells are long (15–35 minutes for 20–50 μM betaxolol), indicative of modulation through slow biochemical cascades. Betaxolol, up to 100μ M, exerted no effects on horizontal cells or the ERG, suggesting that the primary actions of this β₁ blocker are restricted to retinal ganglion cells.

conclusions. These physiological experiments provide supporting evidence that betaxolol acts in a manner consistent with preventing retinal ganglion cell death induced by elevated extracellular glutamate or by increased spontaneous spike rates under pathologic conditions. The physiological actions of betaxolol lead to reducing neurotoxic effects in ganglion cells, which are the most susceptible retinal neurons to glutamate-induced damages under ischemic and glaucomatous conditions. Therefore, betaxolol has the potential to be a neuroprotective agent against retinal degeneration in patients with disorders mediated by such mechanisms.

In the vertebrate retina, glutamate is the primary excitatory neurotransmitter used by photoreceptors, bipolar cells, and ganglion cells. ^{1 2} Under physiological conditions, glutamate activates a variety of receptors, including theα -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), kainate, and N-methyl-d-aspartate (NMDA) receptors, which gate cation channels in the postsynaptic membrane. ^{3 4 5} During pathologic conditions, such as ischemia and elevated intraocular pressure, extracellular glutamate is increased above the normal level, and neurotoxic effects occur in retinal neurons. ^{6 7}  

It has been shown that neurons exposed to elevated extracellular glutamate or ischemia insults exhibit increased spontaneous spike activity. ^{8 9} Sustained high rates of spontaneous spikes may cause cell damage, possibly by allowing excess calcium influx through voltage-gated calcium channels and the NMDA receptor– and perhaps AMPA-kainate receptor–gated channels. ^{10 11} Therefore, substances that decrease spontaneous spike rates and calcium influx through voltage- and glutamate-gated channels are expected to be neuroprotective against cell damage. 

NMDA and AMPA-kainate receptor antagonists have been found to protect neuronal damage caused by elevated glutamate and ischemic insults. ¹² Additionally, a number of substances that do not have direct actions on glutamate receptors also protect neurons against glutamate neurotoxicity. These include opioid receptor antagonists, ¹³ dopamine, ¹⁴ adenosine, ¹⁵ and calcium channel blockers. ¹⁶ Mechanisms underlying interactions between these substances and glutamate receptors are largely unknown. 

A recent report suggests that betaxolol, aβ ₁-adrenergic receptor blocker commonly used in the treatment of glaucoma, may be a retinal neuroprotective agent. Application of betaxolol prevents ischemia-induced reduction in electroretinogram (ERG) b-wave, kainate- or NMDA-induced changes inγ -aminobutyric acid (GABA) immunoreactivity, and intracellular calcium in cultured retinal cells. ¹⁷ However, the report does not provide evidence on what types of retinal neurons are affected by betaxolol and how betaxolol modulates glutamate-induced activities of individual neurons in the retina. Immunocytochemical studies have shown that epinephrine is localized in subpopulations of retinal amacrine cells, which make synapses on ganglion cells. ^{18 19} Therefore, it is possible that retinal ganglion cells contain adrenergic receptors. Moreover, ganglion cells are the only neurons in the retina that exhibit action potentials (spikes), and they are the primary retinal cells that contain NMDA receptors. ^{3 4} These cells are the most susceptible retinal neurons to glutamate-induced damages under ischemic and glaucomatous conditions. ^{20 21} For these reasons, it is important to examine the effects of betaxolol on glutamate-induced spontaneous spike activities, glutamate-induced postsynaptic current, voltage-gated calcium current, and voltage-gated sodium conductance (which is responsible for generating action potentials ²² ) in retinal ganglion cells. Results obtained may provide valuable information on how betaxolol modulates ganglion cell function and how it may protect ganglion cells from glutamate-induced damages. 

Methods

Preparations

Larval tiger salamanders (Ambystoma tigrinum) purchased from Charles E. Sullivan (Nashville, TN) and Kon’s Scientific (Germantown, WI) were used in this study. Single-unit extracellular and intracellular recordings were performed on the isolated, flatmounted retina preparation. During an experiment, the retina and recording electrodes were viewed through a microscope modified for modulation contrast optics (Hoffman Modulation Optics, Greenvale, NY) that was connected to an infrared-sensitive camera and monitor (Cohu, Palo Alto, CA). Animals were dark adapted for a minimum of 1 hour before an experiment. Before the eyes were removed, the animals were first anesthetized by immersing them in 0.04% MS222 and then pithed. The eyes were then enucleated and hemisected. The retina was isolated by pressing the eyecup over a small square of filter paper that had a narrow slit cut out of the center. The filter paper was turned over (so that the retina adhered to the bottom of the filter paper with the ganglion cell layer facing up) and placed in the recording chamber. The recording electrodes could be advanced into the portion of the retina directly underneath the open slit. The preparation rested on two plastic strips located under each side of the filter paper so that the retina was elevated above the glass floor of the recording chamber. This was to prevent the photoreceptors from being pressed against the floor of the recording chamber. To prepare retinal slices, a piece of the posterior half of the eyecup was inverted over a piece of filter (HAO; pore size, 0.45 mm; Millipore, Bedford, MA) secured in the superfusion chamber. The sclera and the pigment epithelium were removed from the retina. The retina and the filter paper were cut into thin slices (approximately 150 mm in thickness) with a custom-made slicer and rotated 90°. ²³ The retinal slices were viewed with a× 40 water-immersion objective lens (Carl Zeiss, Thornwood, NY) modified for the modulation contrast optics. During the experiment, retinal cells and the electrode were clearly observed, and for dark-adapted experiments, a television monitor connected to the infrared image converter (model 4415; Cohu) attached to the microscope was used. 

Electrophysiological Recordings and Solutions

Extracellular recordings were made with platinum-iridium microelectrodes coated with black vinyl lacquer. Intracellular recordings were made with micropipettes drawn with a modified Livingston puller with omega dot tubing (1.0-mm outer diameter and 0.5-mm inner diameter.). The micropipettes were filled with 2 M potassium acetate and had tip resistances measured in Ringer’s solution of 100 to 600 MΩ. The recorded light responses were stored on magnetic tape and subsequently analyzed using a data acquisition system (CED 1401 with Spike2 software; Cambridge Electronic Design, Cambridge, UK). Before data analysis, ganglion cell action potentials were digitized using a window discriminator (Winston Electronics, Millbrae, CA). The retinal preparation was superfused at a rate of 1.0 to 1.5 ml/min with salamander Ringer’s solution (108 mM NaCl, 2.5 mM KCl, 1.2 mM MgCl₂, 2 mM CaCl₂, and 5 mM HEPES, adjusted to pH 7.7 with NaOH) using a peristaltic pump. For the voltage-clamp experiments, patch electrodes of 5-MW tip resistance (series resistance <20 MW) when filled with internal solution containing 118 mM Cs methanesulfonate, 12 mM CsCl, 5 mM EGTA, 0.5 mM CaCl₂, 4 mM ATP, 0.3 mM GTP, 10 mM Tris, adjusted to pH 7.2 with CsOH, were made with patch electrode pullers (Narishige, Greenvale, NY, or David Kopf Instruments, Tujunga, CA). Voltage-clamp recordings were made (Axopatch 200A amplifier connected to a DigiData 1200 interface and pClamp 6.1 software; Axon Instruments, Foster City, CA.). We corrected voltages for the disappearance of the liquid junctional potential at the tips of the patch electrode when the seal was made. Correction varied from −9.2 to −9.6 mV for the electrode internal solution. For simplicity, we corrected all voltage measurements reported in this article by −10 mV. Current responses were analyzed by in-house software and a commercial software package (SigmaPlot; Jandel Scientific, Corte Madera, CA). 

Light Stimulation

A photostimulator with two independent light beams, whose intensity and wavelength could be adjusted by neutral-density filters and interference filters, were used. The light was transmitted to the preparation by way of the epi-illuminator, and the objective lens of the microscope and the spot diameter on the retina could be adjusted by a diaphragm in the epi-illuminator. In most experiments described, large-field illumination (600–1200 mm in diameter) was used. The light source was calibrated with a radiometric detector (United Detector Technologies, Hawthorne, CA) and had an unattenuated irradiance at 500 nm of 5.27 × 10⁶ photons/μm² per second. 

Results

Low Doses of Exogenous or Endogenous Glutamate Increase Spontaneous Spike Activities in Darkness without Blocking Light-Evoked Responses in Retinal Ganglion Cells

We first studied effects of low doses of glutamate in the presence of glutamate transporter blocker d-aspartate on spike activity of retinal ganglion cells. Because retinal cells contain high-affinity uptake transporters that prevent low doses of exogenously applied glutamate from diffusing to the synaptic clefts, ²⁴ coapplication of d-aspartate suppresses such diffusion barriers. ²⁵ Figure 1 shows effects of 50 μM glutamate + 250 μM d-aspartate on the spike activities recorded with extracellular electrode from an ON-OFF ganglion cell in the salamander retina. Spike activities (trace a: spike rates; trace b: spike activity; and trace c: light stimuli[ 0.5-second light steps applied every 10 seconds]) in normal Ringer’s solution (control) are shown in the upper panel, in the presence of 50 μM glutamate + 250 μM d-aspartate are shown in the middle panel, and after drugs were removed (recovery) are shown in the lower panel. Under control conditions, the ganglion cell exhibited a low rate of spontaneous spikes in darkness (0.2–3 Hz) and a high spike rate in response to light stimuli (12–28 Hz). In the presence of 50 μM glutamate + 250 μM d-aspartate, the light-evoked spike rate remained virtually unchanged, whereas the spontaneous spike rate in darkness markedly increased. The spontaneous spikes recovered to the control level after glutamate and d-aspartate were washed out. We obtained similar results with higher glutamate concentration and less d-aspartate (with higher doses of glutamate, glutamate transporters were less effective diffusion barriers). With 200 μM glutamate, no d-aspartate was needed to generate similar results in ganglion cells. However, when we increased the concentration of d-aspartate to 1 mM, no glutamate was needed to generate the similar ganglion cell responses, presumably because when most transporters were blocked, glutamate released from endogenous pools was abundant enough to increase the spontaneous spike activities in retinal ganglion cells. 

Betaxolol Reduces Glutamate-Induced Increase of Spontaneous Spiking in Retinal Ganglion Cells

We next tested the effects of betaxolol on glutamate- and d-aspartate–induced ganglion cell spontaneous spikes. Figure 2 shows that 200 μM glutamate (A) or 1 mM d-aspartate (B) increased the spontaneous ganglion cell spikes from 3 to 10 Hz to approximately 40 Hz, and application of 20 μM betaxolol reduced these increases by approximately 30%. In a sample of seven ON-OFF ganglion cells, the average increase of spontaneous spike induced by 200 μM glutamate was 31.5 ± 6.5 spikes/sec, and that induced by 1 mM d-aspartate was 35.9 ± 5.5 spikes/sec. In the presence of 20 μM betaxolol, the increases became 20.7 ± 3.9 spikes/sec and 23.8 ± 4.6 spikes/sec, respectively. The time-to-action of 20 μM betaxolol varied from 5 minutes to 20 minutes. We detected similar actions of betaxolol at lower concentrations (2–5 μM), but the time-to-action was longer than 30 minutes. 

Betaxolol Reduces Glutamate-Gated Cation Currents in Retinal Ganglion Cells

To determine how betaxolol affects glutamate-induced spontaneous spikes, we examined the actions of betaxolol on glutamate-induced current in ganglion cells under voltage-clamp conditions. Figure 3A shows that puff application of glutamate on a ganglion cell in the retinal slice in Co²⁺ Ringer’s induced postsynaptic currents that reversed near −4 mV. In the presence of 50μ M betaxolol, the glutamate-induced currents were reduced. We obtained similar results in six other ganglion cells. The plot of the average (±SD) difference current-voltage (I-V) relation (I_Control-I_betaxolol) is shown in Figure 3B (filled circles). The difference I-V relation was N-shaped, characteristic of NMDA receptor–mediated current. ⁴ In three other ganglion cells, we recorded the glutamate-induced currents in 50 μM AP5, an NMDA receptor antagonist, and the difference currents (open circles in Fig. 3B ) were very small. These results suggest that the betaxolol-sensitive component of the glutamate-induced postsynaptic current in retinal ganglion cells is mediated predominately by NMDA receptors. 

The time-to-action of 50 μM betaxolol on glutamate-induced current varied from 10 minutes to 25 minutes. We detected similar actions of betaxolol at lower concentrations (5–10 μM), but the time-to-action was longer than 30 minutes. 

Betaxolol Reduces Voltage-Gated Sodium and Calcium Currents in Retinal Ganglion Cells

Results in Figure 2 demonstrate that betaxolol reduced glutamate- or d-aspartate–induced spontaneous spikes in retinal ganglion cells. Because spikes in neurons are mediated by sodium conductance, ²² we therefore studied the effects of betaxolol on voltage-gated sodium currents in retinal ganglion cells. In Figure 4A , we recorded voltage-gated sodium current at various holding potentials under conditions in which the potassium currents were blocked by intracellular cesium in the recording pipette, and the calcium currents were blocked by 1 mM Co²⁺. ²⁶ Application of 50 μM betaxolol reversibly reduced the voltage-gated sodium current to approximately one third its peak amplitude. We obtained similar results from five ganglion cells. The average (±SD) I-V relations of the sodium current in the absence and presence of betaxolol and after betaxolol washout are shown in Figure 4B . The average reduction of peak sodium current by 50 μM betaxolol was 35.7% of the peak control current. 

In view that calcium influx may be a key factor in neurotoxicity, we examined the effects of betaxolol on voltage-gated calcium currents in retinal ganglion cells. In Figure 5A , we recorded voltage-gated calcium current at various holding potentials under conditions in which the potassium currents were blocked by intracellular cesium in the recording pipette, and the sodium currents were blocked by 1 μM tetrodotoxin. ²⁷ Application of 50 μM betaxolol reversibly suppressed the voltage-gated calcium current. We obtained similar results from five ganglion cells. The average (±SD) I-V relations of the calcium current in the absence and presence of betaxolol and after betaxolol washout are shown in Figure 5B . The average reduction of peak calcium current by 50 μM betaxolol was 38% of the peak control current. 

The time-to-action of 50 μM betaxolol on sodium and calcium currents varied from 15 minutes to 25 minutes. We detected similar actions of betaxolol at lower concentrations (2–5 μM), but the time-to-action was longer than 45 minutes. 

Betaxolol Does Not Affect Horizontal Cells or the ERG in the Tiger Salamander Retina

To determine whether betaxolol affects neurons in the outer retina, we examined its actions on horizontal cells (HCs) and the ERG. Figure 6A shows that application of 100 μM betaxolol for 20 minutes (a dose and duration that affect ganglion cells, as described earlier) exerted no effect on either the dark membrane potential or the light-evoked responses of an HC recorded with an intracellular electrode from the flatmounted retina. We obtained similar results from all (six) HCs tested. Figure 6B show that application of 100 μM betaxolol for 20 minutes exerted no significant effect on either the a-wave or the b-wave of the ERG recorded with silver-silver chloride electrode from the eyecup. We obtained similar results from ERGs in all (four) eyecups tested. 

Discussion

Results in this study demonstrate that betaxolol at a concentration of 20 to 50 μM significantly reduced the glutamate– and d-aspartate–induced spontaneous spike activity, the glutamate-induced postsynaptic current, and the voltage-gated sodium and calcium currents in retinal ganglion cells. Neurons exposed to elevated extracellular glutamate exhibit increased spontaneous firing, because glutamate depolarizes the membrane and thus activates the sodium conductance. ⁸ Sustained high rates of spontaneous spikes may cause cell damage, because they repetitively depolarize the cell to positive potentials, allowing excess calcium influx through voltage-gated calcium channels and NMDA receptor–gated channels. ^{11 21} Betaxolol appeared to reduce the glutamate-induced increase of spontaneous spike activities by lowering the sodium conductance (thus raising the action potential threshold) and by suppressing the NMDA receptors. Additionally, betaxolol lowered the calcium conductance in the ganglion cells. We are not certain whether betaxolol interacted directly with the sodium and calcium channels and the NMDA receptors or whether it modulated these molecules through second-messenger pathways. 

It is interesting to recognize that the times-to-action of betaxolol in ganglion cells were very slow. It took 5 to 25 minutes for 20 to 50μ M and more than 30 to 45 minutes for lower doses of betaxolol to affect ganglion cells. For this reason, it is difficult to determine the dose–response relations of betaxolol. Additionally, because of the limited time we could hold cells with electrodes, we could show only reversible actions at higher doses of betaxolol in this study. However, this does not mean that lower doses of betaxolol would have no effects on retinal ganglion cells. In living animals or humans in whom electrode-holding time is not an issue, low concentrations of betaxolol (5 μM or less) may exert similar actions on retinal ganglion cells 30 minutes to perhaps 1 to 2 hours after application. This is even more likely with prolonged administration of betaxolol, as would occur in the treatment of glaucoma. The slow time-to-action of betaxolol nevertheless suggests that this compound may modulate ganglion cell activities through second-messenger systems or other biochemical pathways in retinal ganglion cells. Further studies are needed to clarify this issue. 

Application of 100 μM betaxolol for 20 minutes exerted no action on the HCs and the ERG. The absence of betaxolol action in these experiments was not caused by low dosage or short waiting time, because with the same or lower dosage and waiting time betaxolol substantially affected ganglion cells (see results in Figs. 2 3 4 5 ). Because ERG a-wave is mediated by light responses of the photoreceptors and the b-wave is mediated by bipolar cell and Müller cell responses, ²⁸ our data suggest that betaxolol does not have detectable effects on neurons in the outer retina. Because the primary actions of betaxolol on ganglion cells are modulation of NMDA receptor–gated postsynaptic channels and the voltage-gated sodium channels, and neurons in the outer retina do not contain these two channels, ^{5 29} it is not surprising that we found that betaxolol did not affect the HCs and the ERG. However, our results indicate that betaxolol modulated voltage-gated calcium channels in ganglion cells, and calcium channels have been found in photoreceptors, bipolar cells, and horizontal cells. ^{30 31 32} The absence of betaxolol action on these cells may suggest that ganglion cells and neurons in the outer retina possess different types of calcium channels. Alternatively, it may suggest that calcium channels in outer retinal neurons may not significantly contribute to their dark membrane potential and light-evoked responses. 

It has been suggested that betaxolol is a retinal neuroprotective agent. ¹⁷ Our results support this notion, although we do not show that betaxolol prevented retinal cell death under pathologic conditions. This study demonstrates that betaxolol suppressed the initial physiological changes (increased spontaneous spike rate in ganglion cells) induced by elevated glutamate, and such initial changes are likely to be precursors to excitotoxicity and permanent cell damage. ^{8 12 21} We also show that betaxolol decreased sodium channel conductance, which may lead to higher thresholds for spontaneous spikes and reduced calcium channel conductance, which may suppress calcium influx through these channels. All these physiological actions would lead to reduced neurotoxicity induced by excess spontaneous spikes and calcium influx in ganglion cells. In this study the primary targets of betaxolol actions in the retina were the ganglion cells, and betaxolol did not affect any cells in the outer retina. This is consistent with results from immunocytochemical studies that suggest epinephrine is localized in subpopulations of amacrine cells and that ganglion cells may contain adrenergic receptors. ^{18 19} Although we are not certain that betaxolol modulates ganglion cells through adrenergic receptors, that betaxolol only affected ganglion cells suggests possible links between the two. Moreover, ganglion cells are the most susceptible retinal neurons to glutamate-induced damage under ischemic and glaucomatous conditions, ^{20 21} selective actions of betaxolol on these cells suggest this β₁-blocker may have potential neuroprotective capacities against retinal degeneration in patients with these diseases. 

 Supported in part by grants from Research to Prevent Blindness, Departmental Core Grant P30EY02520 and Grant EY04446 from the National Institutes of Health, the Retina Research Foundation of Houston, and the Neva and Wesley West Foundation of Houston, Texas.

 Presented in part at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, May 1998.

 Submitted for publication February 3, 1999; revised May 17 and September 21, 1999; accepted October 7, 1999.

 Commercial relationships policy: C2, C5, C7 and Cc2, Cc5, Cc7 (RLG).

 

Figure 1.

 

Effects of 50 μM glutamate + 250 μM d-aspartate (D-asp) on the spike activities recorded with an extracellular electrode from an ON-OFF ganglion cell in the salamander retina in normal Ringer’s solution (control; top) in the presence of 50 μM glutamate + 250 μM d-aspartate (middle), and after drugs were removed (recovery; bottom). Trace a: spike rates; trace b: spike activity; and trace c: light stimuli (500 nm −2, 0.5-second light steps applied every 10 seconds).

Figure 2.

 

Effects of 20 μM betaxolol on the increases of spontaneous spike rates induced by 200 μM glutamate (A) or by 1 mM d-aspartate (B) in a ganglion cell. Betaxolol (20 μM) reduced the increases of spontaneous spike activity by approximately 30%. The increase in spike rate induced by glutamate partially recovered about 22 minutes after betaxolol washout (∗), and that induced by d-aspartate recovered about 27 minutes after betaxolol washout (∗∗).

Figure 3.

 

Actions of betaxolol on glutamate-induced current in ganglion cells under voltage-clamp conditions. (A) Postsynaptic currents induced by puff application of glutamate in Co²⁺ Ringer’s at holding potentials −70, −50, −30, −10, 10, 30, and 50 mV. In the presence of 50 μM betaxolol, the glutamate-induced currents were reduced. (B) Average (±SD) difference I-V relation (I_control-I_betaxolol; filled circles) of seven ganglion cells. The difference I-V relation was N-shaped, characteristic of NMDA receptor–mediated current. The average (±SD) difference I-V relation (I_control-I_betaxolol; open circles) of three ganglion cells in 50 μM AP5 are also plotted in (B).

Figure 4.

 

Effects of 50 μM betaxolol on voltage-gated sodium currents in retinal ganglion cells. (A) Sodium current at holding potentials −60, −40, −20, 0, 20, 40, and 60 mV in the three experimental conditions. Potassium currents were blocked by intracellular cesium in the recording pipette, and the calcium currents were blocked by 1 mM Co²⁺. Application of 50 μM betaxolol reversibly reduced the voltage-gated sodium current by approximately one third of its peak amplitude. (B) Average (±SD) I-V relations of the sodium currents of five ganglion cells in the three experimental conditions. The average reduction of peak sodium current by 50 μM betaxolol was 35.7% of the peak control current.

Figure 5.

 

Effects of 50 μM betaxolol on voltage-gated calcium currents in retinal ganglion cells. (A) Calcium current at holding potentials −60, −40, −20, 0, 20, 40, and 60 mV in the three experimental conditions. Potassium currents were blocked by intracellular cesium in the recording pipette, and the sodium currents were blocked by 1 μM tetrodotoxin. Application of 50 μM betaxolol reversibly reduced the calcium current by approximately one third its peak amplitude. (B) Average (±SD) I-V relations of the calcium currents of six ganglion cells in the three experimental conditions. The average reduction of peak calcium current by 50 μM betaxolol was 38% of the peak control current.

Figure 6.

 

Effects of 100 μM betaxolol on an HC recorded with a microelectrode in an isolated retina (A), and the ERG recorded with a silver-silver chloride electrode in an eyecup (B). Application of 100 μM betaxolol for 20 minutes exerted no notable action on HC dark potential, light responses, or a-wave or b-wave of the ERG. The HC was repetitively stimulated by a 0.5-second (650 nm,− 2.3) light step, and the ERG was elicited by 2-second (650 nm, −3.3,− 2.3, and −1.3) light steps.