| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Department of Ophthalmology and Visual Sciences, W. K. Kellogg Eye Center, University of Michigan, Ann Arbor, Michigan; and 2National Institute of Deafness and Communicative Disorders and 3National Eye Institute, National Institutes of Health, Bethesda, Maryland
Submitted 9 April 2004; accepted in final form 19 August 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
). The condition is considered "stationary" because visual function deteriorates minimally with age, unlike the progressive vision loss leading to blindness in retinitis pigmentosa from degeneration and death of the photoreceptor cells (Berson 1993). Two genetic forms of X-linked recessive CSNB have been identified that differ clinically in the extent of impairment of dark-adapted sensitivity of the rod system. The "incomplete-type" CSNB2, at Xp11.23 (Bergen et al. 1996; Rozzo et al. 1999; Strom et al. 1998), from mutations in the gene encoding the alpha-subunit of an L-type calcium channel, retains partial rod-mediated sensitivity. "Complete-type" CSNB1, at Xp11.4 (Boycott et al. 1998; Rozzo et al. 1999), results from mutations in the NYX (nyctalopin) gene that encode a novel protein, nyctalopin (Bech-Hansen et al. 2000; Pusch et al. 2000).
We evaluated retinal function of CSNB1-NYX human subjects. CSNB1 lacks rod-mediated visual perception in darkness, and absolute threshold sensitivity is mediated by cones. Although the function of nyctalopin remains unknown, electrophysiology studies of CSNB1 using noninvasive electroretinogram (ERG) recordings indicate that the defect lies postsynaptic to the rod photoreceptors, as the rod-driven a-wave remains normal but the dark-adapted ERG b-wave is absent (Miyake et al. 1986). The a-wave originates from hyperpolarization of the rod photoreceptors on exposure to light stimuli (Baylor et al. 1984; Granit 1947; Penn and Hagins 1972), whereas the b-wave reflects activity of postsynaptic bipolar cells directly or indirectly (Gurevich and Slaughter 1993; Robson and Frishman 1995; Xu and Karwoski 1994). The rod pathology has only depolarizing bipolar cells (DBC) (Dacheux and Raviola 1986; Wassle and Boycott 1991).
Signaling through the cone pathways is also defective in CSNB1, as the light-adapted ERG b-wave is abnormal in CSNB under photopic conditions with brief photostrobe stimuli (Heckenlively et al. 1983; Hill et al. 1974; Miyake et al. 1986; Quigley et al. 1996; Young 1991). Localizing this deficit on a cellular/pathway basis is difficult, however, as the cone retinal circuitry involves parallel signaling through DBCs of the cone ON pathway as well as through hyperpolarizing bipolar cells (HBCs) of the cone OFF pathway (Sterling et al. 1992). Although activity of cone-driven DBCs is requisite for generation of the photopic b-wave, cone HBC activity modulates and shapes the response (Sieving et al. 1994), making it difficult to parcel out the extent of ON/OFF-pathway involvement in CSNB1 with photostrobe flashes.
Cloning the NYX gene has not resolved this issue, and the cellular mechanism by which nyctalopin causes human night blindness remains uncertain. Nyctalopin encodes a small proteoglycan (Bech-Hansen et al. 2000; Pusch et al. 2000) with 11 leucine-rich repeats (LRRs) that are important for protein–protein interactions (Hocking et al. 1998). Human and mouse nyctalopin are membrane-bound extracellular proteins that localize over the cell surface (Zeitz et al. 2003). Other LRR family members are involved in cell adhesion and axon guidance (Pusch et al. 2000). In situ hybridization with a 668-bp NYX antisense probe showed labeling throughout several retinal layers and hence did not localize the functional defect (Bech-Hansen et al. 2000). Nyctalopin is abundant in the retinal ganglion cell and amacrine cell layers of mouse and rat (Pesch et al. 2003), but neither ganglion cell nor amacrine cell dysfunction would cause the ERG b-wave abnormality. The question remains as to whether only the ON-pathway circuit is involved and at what level (Bech-Hansen et al. 2000; Pesch et al. 2003).
In this study, we used noninvasive ERG recordings to evaluate NYX genotyped CSNB1 males and then correlated this with ERGs recorded identically from non-human monkey primates during intravitreal application of glutamate analogues that suppress light-evoked signaling of depolarizing or hyperpolarizing bipolar cells selectively. Photopic ERG responses were elicited with sinusoidal and ramping rapid-ON/OFF flicker stimuli that are proposed to probe the cone ON and OFF pathways in semi-selective fashion (Alexander et al. 2001b; Kremers et al. 1993; Roveri et al. 1997), although this latter assertion has not actually been put to robust analysis for ERG studies in either human or monkey. The results localized the CSNB1-NYX photopic functional deficits to the DBC pathway with no apparent involvement of HBCs. The data also gave insight into the limits of using ramping ON/OFF ERG stimuli to probe retinal ON versus OFF pathways.
This work was presented in part at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, FL, May 2001 and May 2004.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Informed consent was obtained from all subjects under protocols approved by the University of Michigan Medical School Institutional Review Board. Four male subjects, 17–20 yr old, were studied from three CSNB1 pedigrees. Six clinically normal male and female volunteers, ages 33–46 yr, were control subjects. Best corrected visual acuity of the CSNB1 subjects was 20/30 to 20/60, except for the right eye of one subject with 20/200 acuity due to amblyopia. Control subjects had 20/20 corrected acuity. CSNB1 subjects frequently are myopic (Allen et al. 2003; Heckenlively et al. 1983; Hill et al. 1974; Jacobi et al. 2002; Quigley et al. 1996), and their refractive error range was –6.5 to –14.00 diopters (D). The controls ranged from –3.5 to –13.00 D. Color vision was normal by Ishihara plate testing for all subjects. Visual fields were full or minimally altered for all subjects with the Goldmann perimetry V4e and I4e targets. Dark-adapted final thresholds were determined after 45 min in the dark with a Goldmann-Weekers Dark-Adaptometer. All four CSNB1 subjects had 2.5–3 log elevation of threshold and complete loss of rod-mediated sensitivity. Control subjects had normal thresholds. Fundus appearance of the CSNB1 subjects was myopic but otherwise unremarkable except for slight pigmentary granularity of the retinal pigment epithelium as can be found in some CSNB individuals (Sieving 1993).
Genotyping and mutation analysis
DNA was isolated from peripheral blood lymphocytes, and the three NYX exons (Bech-Hansen et al. 2000; Pusch et al. 2000) and flanking intron-exon sequences were amplified using PCR primers (Table 1) designed from the database sequences of clone Z9301 for this region (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=nucleotideandcmd=searchandterm=Z93015).
|
; Bech-Hansen et al. 2000; Dunnen and Antonarakis 2000; Pusch et al. 2000). Computational analysis of the mutations and the putative effects on the protein translation and structure was performed using DNASTAR (Madison, WI), GCG (Madison, WI), and ExPASy molecular biology server (www.expasy.ch/). Human ERG recording
Pupils were dilated fully with 1% tropicamide and 10% phenylephrine HCl, and 1% proparacaine corneal topical eye drops were administered for anesthesia prior to placing ERG recording electrodes (bipolar Gold lens electrodes, Doran Instruments, Littleton, MA). A neutral recording skin electrode was taped to the forehead. CSNB1 clinical ERG diagnostics were performed according to ISCEV standards (Marmor and Zrenner 1998) using a Ganzfeld full-field configuration for the brief xenon flash stimulus. Scotopic recordings were performed first, after an initial 45-min dark-adaptation period, using dim blue stimuli (Kodak Wratten 47; –1.82 log cd-s/m2 per flash) and brighter white stimuli (0.83 log cd-s/m2 per flash). Subjects were then light adapted for 10 min by exposure to a white 42 cd/m2 rod saturating background, and photopic ERGs were recorded for single flash white stimuli (0.83 log cd-s/m2 per flash) and to white 32 Hz strobe flicker trains (Sieving et al. 1998) at 0.46 log cd-s/m2 per flash.
Photopic ERG responses were also elicited with sinusoidal and ramping rapid-ON/OFF-flicker stimuli. "Rapid-ON" signifies an abrupt increase in stimulus luminance followed by a ramping decline back to baseline. These stimuli were generated by modulating a densely packed array of 91 red LEDs (623-nm wavelength, 8-nm half-width). The 65-mm-diam LED array was placed at the top of an 8-in-diam integrating bowl with a small viewing port 90° away on the side that was positioned directly in front of the eye to provide a full-field Ganzfeld stimulus. The LED light intensity was controlled by current modulation by a digital function generator (Wavetek Model 39, San Diego, CA) at temporal frequencies of 4–56 Hz. Light output determined as a function of input voltage was found to be highly linear at the light levels used in this study. Both the sinusoidal and the ramping stimuli had 480 cd/m2 maximum and 24 cd/m2 minimum stimulus intensities superimposed on a 42 cd/m2 constant white background, giving mean luminance of 294 cd/m2 with 78% modulation depth. Rise time of the rapid-ON stimulus was 5.2 µs, and the fall time of rapid-OFF stimulus was 4.7 µs. The rise or fall time was 
0.03% of the period at the highest temporal frequency of 56 Hz, and at longer periods, this is even more negligible. The stimulus ramp occupied the remaining portion of each cycle.
Subjects remained light-adapted and were preadapted to the stimulus for 5 s at each stimulus frequency. Data were then collected for 10 s for sinusoidal and for rapid-ON- and rapid-OFF-flicker stimuli. Two to four data sets were averaged off-line. Responses were amplified at 10,000 gain from 1 to 1,000 Hz and digitized at 3,072 Hz rate. Amplitude and phase of the fundamental component and higher harmonics of responses to sinusoidal stimuli were derived by Fourier transform analysis (MatLab Signal Processing Toolbox, The MathWorks, Natick, MA).
Monkey ERG recordings and intravitreal drug application
Rhesus monkeys (Macaca mulata) were studied under protocols approved by the Unit of Laboratory Animal Medicine at the University of Michigan. The animals were sedated with ketamine hydrochloride (7 mg/kg im; 5 to 10 mg · kg–1 · h–1 im maintenance dose) and xylazine (0.6 mg/kg im). Supplemental oxygen was given by external nasal cannula. Respiration and heart rate were monitored. Hydration was maintained by subcutaneous lactated Ringer solution. Pupils were dilated with topical 1% atropine and 2.5% phenylephrine HCl. ERGs were recorded with Burian-Allen bipolar corneal contact lens electrodes (Hansen Ophthalmic Development Labs, Coralville, IA) after corneal topical anesthesia (0.5% proparacaine) and lubrication with methylcellulose. The ERG responses to sinusoidal and ramp flicker stimuli were elicited from monkey using the same equipment and the same recording conditions as for the human data.
Intravitreal application of dl-2-amino-4-phosphonobutyric acid (APB) (Sigma Chemical) was given to suppress the activity of depolarizing bipolar cells (Shiells et al. 1981; Slaughter and Miller 1981), and cis-2,3-piperidine-dicarboxylic acid (PDA) (Sigma Chemical) was given to suppress hyperpolarizing cell activity (Sieving et al. 1994; Slaughter and Miller 1983). Combining both drugs eliminates essentially all glutamate-mediated signaling from all classes of bipolar cells. Drugs were freshly dissolved in sterile saline 2 h before injection, adjusted to 5.5–7.0 pH using HCl or NaOH for solubility and passed through a 0.45-µm filter. A 30-gauge needle was inserted through the pars plana 6-mm posterior to the limbus, and intravitreal drug injection volumes of 0.03–0.05 ml were administered. This yielded diluted vitreal concentrations of 2 mM APB and 4 mM PDA in the monkey, assuming full dilution by the vitreal volumes (Knapp and Schiller 1984). Monitoring of ERG waveforms indicated full drug effects within 1 h after injection.
Five monkeys not previously studied were employed for ERG recordings, and the sinusoidal flicker results were augmented by data from three monkeys that we partially described previously (Kondo and Sieving 2001). The rapid-ON/OFF-flicker data were generated from the five newly studied monkeys. The effects of APB followed by PDA were determined in two monkeys, and the effects of PDA followed by APB were recorded in three monkeys.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
NYX sequence alterations were found for all four CSNB1 male subjects. Family A: two affected brothers, subjects S1 and S2, had a novel G to C transversion in exon-3 (803G 
C) causing a missense codon change R268P within the leucine rich repeat LRR-9 motif. Secondary structure changes are predicted to alter the turns and coils in the region and local amphipathicity. Family B: subject S3 had a double nucleotide change (710T A and 711C A) in exon 3 causing a missense L237Q in the LRR-8 motif. These mutations are previously unreported. Family C: subject S4 had a 24 nucleotide deletion (85–108 del 24 nt) in exon-3 near the protein N-terminus causing an in-frame deletion of eight amino acid residues (29–36) of the N-terminal cysteine cluster. The same 24-bp deletion has been reported previously for seven other apparently independent U. S. families (Bech-Hansen et al. 2000).
ERG responses to pulse stimuli
None of the four subjects showed rod-mediated threshold sensitivity on clinical examination. Their dark-adapted ERG responses retained normal a-wave amplitudes but essentially lacked the b-wave, which was reduced to only 6% of normal mean scotopic b-wave amplitude. As the dark-adapted response depends primarily on activity of the rod DBC pathway, this indicates impaired signaling of rod driven depolarizing ON bipolar cells. The photopic ERG showed the broad a-wave trough that is characteristic of complete-type CSNB (Sieving 1993). Photopic ERG b-wave responses of CSNB1-NYX subjects to longer 150-ms duration stimuli had major reductions in the photopic b-wave, but the d-wave persisted (Fig. 1). This b-wave loss results from deficient activity of the postphotoreceptoral ON pathway (Sieving et al. 1994). Full analysis of the d-wave, however, is not firmly established, and attribution of d-wave responses exclusively to the OFF pathway is tenuous as the following results demonstrate.
|
ERG responses to sinusoidal stimuli of CSNB1-NYX subjects were different from controls across a range of stimulus frequencies (Fig. 2). The CSNB1-NYX waveforms were more sinusoidal than controls across stimulus frequencies of 4–56 Hz and resembled the monkey waveforms after either PDA (Fig. 3) or APB (e.g., Fig. 2 in Kondo and Sieving 2001), which simplify the signaling to a single retinal pathway.
|
|
). Applying APB and PDA together to block inner retinal activity showed that the photoreceptor component amplitude declined as stimulus frequency increased (Fig. 3).
Figure 4 shows the amplitude and phase of monkey and human responses to sinusoidal stimuli. Normal human and control monkeys without drug application have similar response profiles across stimulus frequency with maximum amplitude between 32 and 40 Hz, and an amplitude dip and phase inflection near 12 Hz as noted previously in human (Alexander et al. 2000; Burns et al. 1992; Odom et al. 1992) and in non-human primates (Kondo and Sieving 2001). CSNB1-NYX responses lack both the amplitude dip and the phase inflection near 12 Hz. Intravitreal application of either APB or PDA in monkey eliminated the amplitude dip and the phase inflection, indicating that both pathways contribute to the net response and that the phase inflection results from an interaction between the pathways. Application of APB in monkey to eliminate ON-pathway signaling replicated the absolute phase relationship of CSNB1-NYX relative to human controls, whereas the absolute phase after PDA was quite different from CSNB1-NYX.
|
; Odom et al. 1992), and its amplitude was comparable to the fundamental amplitude. After APB or PDA, the amplitude of the fundamental component increased (Fig. 4), but the amplitude of the second harmonic increased only slightly at some frequencies. The postdrug behavior of the second harmonic roughly mimicked in CSNB1-NYX, but further studies and detailed analyses will be required to fully determine the effects. Flicker ERG to rapid-ON/OFF stimuli
CONTROL RESPONSES.
The ERG was further evaluated using photopic rapid-ON and -OFF ramping stimuli (Fig. 5). In single-unit recordings of monkey retinal ganglion cell activity, these stimuli preferentially stimulate ON- and OFF-pathway activity respectively (Kremers et al. 1993). Rapid-ON/OFF ramp stimuli have been used clinically to probe the relative ON- versus OFF-pathway deficits in X-linked retinoschisis by ERG recordings (Alexander et al. 2001b), in melanoma-associated retinopathy (Alexander et al. 2002), in carriers of X-linked retinitis pigmentosa (Alexander et al. 2003), and in human psychophysical studies to investigate the sensitivity to increments and decrements of light (Bowen et al. 1989; Purkiss et al. 2001).
|
32 Hz, the intervals were sufficiently short that the components merged for both normals and CSNB1-NYX, and discrete components were difficult to judge. By comparison, with rapid-OFF stimuli, the CSNB1-NYX subjects had responses quite similar to normals although slightly larger; the positive response to rapid-OFF stimuli corresponded to the d-wave observed at termination of a step stimulus (Fig. 1). Monkey responses to rapid-on and rapid-OFF ramp stimuli were dissected into constituent pathway components by drug application (Fig. 6). Responses after either APB or PDA alone were larger than control for at least some stimulus frequencies, indicating that both pathways contribute to control responses as would readily be suspected from the waveform alone at 4 Hz.
|
). One can conclude that the d-wave is a complex response that contains contributions from both pathways and does not derive exclusively from the OFF pathway. Drug-sensitive components to rapid-ON/OFF stimuli
ON- and OFF-pathway activity was isolated by waveform subtractions after drug application in monkey (Gurevich and Slaughter 1993; Hood et al. 2002; Robson and Frishman 1995; Sieving et al. 1994) and then compared with the CSNB1 subjects. The APB-sensitive component in Fig. 7 can be obtained in two ways: first by subtracting the post-APB records from the control records (control –APB) or second, by applying APB after first applying PDA [i.e., PDA –(PDA+APB)]. Both conditions highlight the ON-pathway ERG contribution, but the physiological conditions are different. In the (control –APB) condition, the OFF pathway remains active, even though the process of subtraction removes this component from view. In the PDA –(PDA+APB) condition, the OFF pathway is pharmacologically suppressed. This analysis presumes that the ERG represents three major components: activity of the ON pathway, the OFF pathway, and photoreceptors. For all waveforms in Fig. 7, the subtractions remove the photoreceptor component from view. Permutations of this scheme highlight the PDA- and APB-sensitive components.
|
). We recorded these responses across a range of stimulus frequencies, but we could not condense all of this information by harmonic analysis because Fourier series approximation for these responses would not be appropriate particularly at the longer stimulus intervals in which the responses are isolated periodic events rather than being continuous. Consequently, we compared human and monkey responses only for 12-Hz ramping stimuli that are near the stimulus frequency that elicits the amplitude dip and phase lag inflection for sinusoidal stimuli. The major ERG components are still distinguishable in the waveforms elicited at 12 Hz. Although we could not illustrate the entire data set across all stimulus frequencies, the major themes of the results that are shown after drug isolation are similar for other frequencies. The central conclusion is that rapid-ON/OFF ramping stimuli elicit complex responses that have major contributions from both the ON and OFF pathways. One simply cannot equate responses to rapid-ON ramp stimuli exclusively with the ON-pathway or OFF-pathway activity with rapid-OFF ramp stimuli. Comparison of CSNB1 and monkey responses
Finally, we compared responses of a representative human CSNB1-NYX to drug-isolated monkey components (Fig. 8). For reference, 1 shows that normal monkey and normal human control responses (normalized for peak amplitudes) are quite similar for both stimuli.
|
Suppressing DBC signaling in monkey with intravitreal APB yields a waveform with great fidelity to CSNB1-NYX, for both rapid-on and -off stimuli (Fig. 8, 3), indicating that the ON pathway contributes very little to the CSNB1 response. These waveforms after APB (in 3) represent approximately intact retinal activity except that the ON pathway is suppressed, i.e., the contributions are primarily from OFF-pathway activity plus photoreceptors. The same conclusion, that CSNB1-NYX responses reflect primarily OFF-pathway activity, can be reached by looking at the PDA-sensitive component in Fig. 8, 4. Again, the leading edge of the PDA-sensitive component (plus any direct cone contribution) corresponds quite well with CSNB1, indicating that OFF-pathway activity is sufficient to explain the CSNB1-NYX waveform.
Finally, suppressing the PDA-sensitive OFF-pathway signaling in monkey yields a waveform quite dissimilar to CSNB1-NYX (Fig. 8, 5), further indicating the necessary presence of HBC signaling contributions to CSNB1-NYX responses.
Fig. 8, 5, also captures the negative-going waveform to rapid-OFF stimuli contributed by the ON-pathway system. The genetic elimination of this negative-going activity from CSNB1-NYX responses explains the larger than normal amplitude seen in Fig. 8, 2 rapid-OFF and in the waveforms to termination of the 150-ms stimulus in Fig. 1.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
None of the four CSNB1 males showed the response amplitude dip near 12 Hz that normal controls exhibited to sinusoidal stimuli (Alexander et al. 2000; Burns et al. 1992; Odom et al. 1992). In non-human primate, the amplitude dip results from phase cancellation between ON- and OFF-pathway components (Kondo and Sieving 2001). This phase inflection requires joint activity of ON and OFF pathways together and was not evident in the isolated activity of either pathway alone, as application of either APB or PDA eliminated the phase reversal in monkey. CSNB1-NYX males also did not show the normal phase inflection with 12-Hz stimuli. One can suspect that even partial activity of both pathways might preserve at least some degree of phase reversal, suggesting that the ON pathway is completely suppressed in these four CSNB1-NYX subjects. One might evaluate this further through study of incomplete-type CNSB subjects who show reduced d-wave amplitudes (Allen et al. 2003).
These results do not indicate the state of DBC membrane polarization in CSNB1-NYX. APB is a noncompetitive glutamate agonist that activates the sign-inverting metabotropic glutamate DBC receptor and thereby mimics constant darkness in which rods and cones continuously release glutamate. Darkness and APB both hyperpolarize the DBCs. However, mimicking CSNB1 with APB only gives evidence of impaired DBC signaling but not the DBC polarization state. The present data are uninformative regarding the state of HzCs in CSNB1-NYX. However, as HzC feedback is expected to modulate photoreceptor responses and the a-wave remains normal in most CSNB1-NYX (Allen et al. 2003), the presumption is that HzC signaling is not affected.
The ERG of CSNB1-NYX was quite different from the cone photoreceptor responses isolated in monkey by applying APB and PDA simultaneously. The cone responses were small and exhibited low-pass frequency characteristics in which the amplitude decreased monotonically for stimulus frequencies >4 or 8 Hz, for sinusoidal as well as for rapid-ON-OFF ramping stimuli. Although Burns et al. (1992) attributed the fall-off of high-frequency response to an early temporal filter within the cone photoreceptors, the present data and our past results with flicker ERG in monkey indicate that the cones directly contribute very little >16 Hz. Hence it is unlikely that much of the CSNB1-NYX amplitude reduction at higher frequencies results from a lack of photoreceptor amplitude. It is possible, however, that APB and PDA may also have effects elsewhere in the retina including horizontal cell feedback onto photoreceptors in some species (Massey 1995; Nawy 1989; Yang 1989). Blocking horizontal cell feedback could reduce photoreceptor flicker amplitude in vivo.
NYX mutations in these three families
All four CSNB1-NYX males had similar ERG responses despite three different mutations in the NYX gene. The mutations in all three families involved the leucine-rich repeat LRR regions of the protein in structurally conserved residues of the gene product that are predicted to coincide with a considerable neuro-signaling deficit (Jacobi et al. 2002). All four CSNB1-NYX affected males had high myopia (–6.5 to –14.00 D) and had subnormal ERG b:a-ratios of 0.45–0.88 (normally >1.0), similar to findings of 12 CSNB1 males with NYX mutations in conserved domains (Jacobi et al. 2002). A majority of the missense mutations reported previously also localized to LRR regions: LRRs NT, 4, 5, 6, 11, CT (Pusch et al. 2000) and LRRs NT, 4, 6, 8, 9, 10 (Bech-Hansen et al. 2000), indicating a critical role of these motifs in nyctalopin function.
Possible residual rod pathway function in some CSNB1
While our CSNB1-NYX subjects did not appear to have residual rod pathway function, evidence from other studies suggests that although greatly impaired, some residual rod system function may remain for some CSNB1-NYX males (Allen et al. 2003; Bech-Hansen et al. 2000; Scholl et al. 2001), possibly through an alternate rod pathway. Mammalian rods contact a single type of bipolar cell (Boycott and Wassle 1991; Dacheux and Raviola 1986; Dowling and Boycott 1966), but a second pathway for rod signals has been suggested through gap junctions onto cones and then to the cone bipolar cells (DeVries 1995; Nelson 1977; Schneeweis and Schnaff 1995; Smith 1986). Recent investigations have shown a possible third pathway for rod signals, wherein rod photoreceptors bypass the rod bipolar cell, and directly excite OFF-cone bipolar cells in cat (Fyk-Kolodziej et al. 2003) and mouse (Hack et al. 1999; Soucy et al. 1998).
Mendelian human traits as model systems for retinal function studies
We still do not have full understanding as to how selective impairment of retinal pathway activity will affect human ERG responses. Knowledge of DBC and HBC contributions to the primate ERG has come primarily from invasive studies of non-human primate retina by injecting glutamate analogues to suppress the activity of one or another class of bipolar cells. This approach cannot be used routinely to study human retinal function in vivo, although a few in vitro ERG studies applied aspartate to human eyes enucleated for medical reasons (Yonemura et al. 1974). DBC and HBC contributions can be evaluated in rodent and amphibian retinas, (Awatramani et al. 2001; Xu and Karwoski 1994, 1995; Xu et al. 2003) although these results probably do not translate to human primate with much fidelity, particularly for the cone system that shows considerable variation across species (Evers and Gouras 1986)
Non-human species also exhibit genetic retinal dystrophies and can provide additional information. The naturally occurring X-linked nob ("no b-wave") mouse, with an 85-bp deletion in the murine NYX gene (Gregg et al. 2003), has ERG features of rod pathway deficits similar to human CSNB1 subjects (Pardue et al. 1998). The mouse retina has the major classes of DBC and HBC bipolar cells, and it could be useful for understanding the human trait. Unfortunately the photopic OFF-ERG response, that is characteristic of retinal I-type ("inhibitory-type") responses in the primate retina is quite weak in mice (Evers and Gouras 1986). Temporal frequency response function to sinusoidal flicker in mouse differs from the primate and human response functions in that it does not exhibit the amplitude minimum near 12 Hz and the maximum between 32 to 40 Hz (Krishna et al. 2002). Moreover, the nob mouse, has a temporal frequency response that is quite unlike that of CSNB1 and post-APB monkey response (Krishna et al. 2002).
There is increasing evidence that different bipolar cell types are tuned to different temporal frequencies. In cat, bipolar cell types b1–b3 connect via parallel circuits from cones to the ON-
(X) ganglion cell (Cohen and Sterling 1992; McGuire et al. 1986). All three types of cells connect to the same set of cones and carry the same spatial and spectral information but convey different temporal information (Cohen and Sterling 1992). Freed (2000) isolated bipolar cell responses from ganglion cell responses to a step of light with tetrodotoxin in cat and showed that the b1 bipolar cell contributed the high-frequency response, whereas b2 and b3 bipolars contributed the lower temporal frequencies. In context of the present study, responses >30 Hz for both sinusoidal and ramping rapid-ON/OFF stimuli were similar in controls and CSNB1-NYX as well as after APB in monkey. One could speculate that nyctalopin acts mostly on the b2 and b3 bipolar cells and less on the b1 cell.
Central vision in CSNB1-NYX
CSNB1-NYX individuals essentially never reach normal acuity (Heckenlively et al. 1983; Jacobi et al. 2002; Krill and Martin 1971; Miyake et al. 1986; Young et al. 1986), indicating that ON-pathway processing is required to take full visual advantage of spatial sampling by the foveal cone photoreceptor matrix. Although unlikely, one must ask whether foveal cone spacing is altered in CNSB1.
Color vision in CSNB1-NYX is surprisingly good despite the cone ON-pathway dysfunction. Short-wavelength-sensitive cone (S-cone) ERGs are nonrecordable to full-field stimuli in CSNB (Kamiyama et al. 1996). Recently, blue-on-yellow perimetry in CSNB1 patients, however, showed normal blue cone sensitivity in the central retina, but this became abnormal in the periphery (Terasaki et al. 1999). Anatomical evidence indicates that primate foveal S cones contact ON blue bipolar cells (Klug et al. 2003; Kouyama and Marshak 1992; Mariani 1984). Unlike the L or M cones, however, the S cones are not presynaptic to their own depolarizing midget bipolar cells. Rather, S cones contact hyperpolarizing midget bipolar cells (Ahmad et al. 2003; Klug et al. 2003; Kolb et al. 1997), which provide chromatic input to ganglion cells in both the fovea and periphery (Boycott and Wassle 1991). Blocking only the ON pathway with APB did not change the shape of the spectral sensitivity function, but it reduced sensitivity uniformly across all wavelengths, suggesting that the OFF pathway can carry short-wavelength information (Smith et al. 1989).
As the full-field ERG cannot detect abnormalities limited to the fovea, one could speculate that the NYX proteoglycan molecule may have no affect in the central foveola. Alternatively, there may be connections/circuits that support good color vision without a postreceptoral cone ON pathway in the central retina. Focal macular cone ERGs recorded from CSNB1 patients (Kondo, unpublished observation) had waveforms quite similar to APB-treated monkeys, suggesting that the cone ON pathway may be blocked completely even in the central retina in these patients. Obviously these issues are not yet resolved.
Rapid-ON/OFF stimuli elicit complex primate ERG origins
It would be very helpful clinically to have a noninvasive ERG tool to probe retinal ON- and OFF-pathway integrity. Photopic "b-waves" that are elicited by instantaneous photostrobe stimuli that are traditionally used for clinical studies are a composite of ON- and OFF-ERG responses (Sieving 1993). Preservation of either ON- or OFF-pathway activity will leave this photopic photostrobe b-wave response at least partially intact, and hence photostrobe ERGs are not fully informative regarding the specific deficits found in CSNB1 responses.
Rapid-ON/OFF ramping stimuli have been used in psychophysical (Bowen et al. 1989; Purkiss et al. 2001) and in physiological primate studies (Kremers et al. 1993) to separate the two pathways. Some have used these stimuli to explore pathway deficits in several human hereditary retinal dystrophies (Alexander et al. 2001b, 2002, 2003) with the expectation that the rapid-ON/OFF stimulus paradigms are relatively selective for the respective retinal physiological ON/OFF pathways. This possibility was not tested experimentally, however, prior to the ERG data presented here. This is the first study of primate ERGs to rapid-ON/OFF ramping stimuli during pharmacologic separation of ON- and OFF-pathway activity. To our disappointment, the data show that these stimuli are nonspecific for pathway separation and yield complex ERG responses in primate. Rapid-ON stimuli elicit considerable activity of both the ON and OFF pathways, as does the rapid-OFF stimulus. Neither is truly selective for a single pathway.
The caveat, however, is that pharmacological dissection may not yield unequivocally complete isolation of components. This seems to be less of a concern for APB results than with PDA, which can also act on horizontal cells and third-order neurons of some species (Slaughter and Miller 1985). APB is a highly selective competitive antagonist at the mGluR6 receptor that localizes exclusively at the metabotropic G-protein-coupled synapse between rod and cone photoreceptors and their respective DBCs (Shiells et al. 1981; Slaughter and Miller 1981). APB does not act at the chemical synapse onto HBCs, although it may alter horizontal feedback onto photoreceptors in some preparations (Hare and Owen 1992; Nawy et al. 1989; Takahashi and Copenhagen 1992; Yang and Wu 1989).
Hence, it is important to note the remarkable convergence of results of pharmacologic manipulation in monkey and genetic alteration in these four CSNB1-NYX subjects. CSNB1 individuals apparently afford the opportunity to probe the consequences to visual function of eliminating ON-pathway signaling in the human primate retina.
These ERG findings with rapid-ON/OFF stimuli extend previous results for brief strobe flicker and 30-Hz sinewave (Bush and Sieving 1996; Kim et al. 1997; Kondo and Sieving 2001, 2002). In all cases, the photopic ERG waveforms showed major postphotoreceptoral contributions. The results presented here with sawtooth-modulated flicker stimuli similarly are largely from the inner retina, particularly for stimuli >16 Hz. Further development of ERG analyses is warranted, to determine the extent to which attribution to a specific pathway might be extracted from such waveforms.
Third-order neuron signaling in CSNB1-NYX
In situ study showed NYX expression across all retinal layers including amacrine cells of the inner nuclear layer (Bech-Hansen et al. 2000), raising a question of whether ERG abnormalities might be attributed to deficient signaling by third-order neurons. Retinal amacrine and ganglion cell neurons do not contribute substantially to sinusoidal flicker responses in monkey, and the contributions to the fundamental component are small in comparison to those from bipolar cells (Viswanathan et al. 2002). Blocking inner retinal activity of ganglion and amacrine cells with TTX and N-methyl-D-aspartate shifted the amplitude dip in the response fundamental to a lower stimulus frequency and the maximum to higher stimulus frequency (Viswanathan et al. 2002). This is dissimilar to our sinusoidal stimuli results with CSNB1-NYX subjects, in which the amplitude dip was abolished and the maximum remained near 30 Hz. In the least, this reinforces that the cellular locus of dysfunction likely involves a site other than or in addition to third-order neurons.
Several ERG studies of CSNB suggested that signaling is deficient through the proximal retina. Miyake et al. (1994) interpreted the absence of a scotopic threshold response (STR) in complete-type CSNB subjects as indicating a disturbance in the proximal rod neural pathway. The STR (Sieving et al. 1986) originates in the retina proximal to bipolar cells, through activity of A2 and A17 amacrine cells (Naarendorp and Sieving 1991). The STR is uninformative for CSNB, however, as a signaling defect upstream of DBCs would suppress signals from reaching the amacrine cells and generating an STR. Photopic oscillatory potentials (OPs) are also reduced or even absent in CSNB (Allen et al. 2003; Heckenlively et al. 1983; Lachapelle et al. 1998; Tremblay et al. 1995). As the OPs originate downstream of the DBCs, suppressing DBC signaling with APB must affect the OPs, and one cannot distinguish a primary effect on third-order neuron signaling. In monkey (Ueno et al. 2004) and the amphibian tiger salamander (Awatramani et al. 2001), third-order retinal neurons influence the b- and d-waves, but again this signaling originates downstream of the DBC deficit in CSNB1 and is uninformative.
Limitations of functional dissection with the clinical ERG
Analysis of the cone system is particularly difficult using the ERG because parallel signaling occurs through the DBC ON pathway and HBC OFF pathway. Both systems contribute to the photopic ERG response elicited by photostrobe stimuli traditionally used for clinical studies and in one recent CSNB1 analysis (Jacobi et al. 2002). The photostrobe photopic "b-wave" is a concatenation of the true b- and d-waves evoked with longer stimuli (Sieving 1993). Preservation of either ON- or OFF-pathway activity will leave the photopic "b-wave" to photostrobe stimuli at least partially intact and hence is not fully informative regarding the specific deficits found in CSNB1 responses.
Longer-duration photopic stimuli provide additional information by temporally separating the b- and d-waves. The long-flash photopic b-wave at light onset is attenuated in CNSB, whereas the d-wave at light termination retains substantial amplitude (Allen et al. 2003; Holopigian et al. 1991; Houchin et al. 1991; Miyake et al. 1987; Quigley et al. 1996; Sieving 1993). The b-wave under long-flash conditions requires DBC activity (Stockton and Slaughter 1989; Tian and Slaughter 1995), although HBC OFF-pathway activity exerts substantial control over b-wave amplitude and waveform (Sieving et al. 1994). Consequently, when the b-wave is abnormal and has either greater or lesser amplitude or abnormal timing, one cannot attribute this exclusively to an ON-pathway abnormality. Some recent literature implies that one can equate the photopic b-wave to long-flash stimuli with the ON pathway and the d-wave with the OFF pathway (Scholl et al. 2001). This is an oversimplification of a complex ERG response.
Speculations on the functional role of nyctalopin
In situ hybridization with a 668-bp antisense probe corresponding to the NYX 3' UTR region showed labeling of several retinal layers, including the inner layers and retinal ganglion cells (Bech-Hansen et al. 2000). Immunohistochemical studies in the nob mouse to examine the distribution of proteins in the outer plexiform layer that have been implicated in DBC function showed no abnormalities in the rod bipolar cell dendrite morphology (Ball et al. 2003), indicating that the absence of nyctalopin does not disrupt the expression pattern of other proteins that are necessary for synaptic transmission. NYX is also expressed in kidney (Bech-Hansen et al. 2000) and possibly other tissues (Pusch et al. 2000), although human CSNB1 subjects are not known to have renal or other systemic abnormalities. Nyctalopin expression in the nob mouse carrying a murine NYX mutation was greatest in the inner nuclear layer that contains bipolar, horizontal and amacrine cells (Gregg et al. 2003). Consequently these studies provide few clues as to specific ON/OFF-pathway deficits.
NYX in situ analysis suggested that nyctalopin might play a role in neural retinal development (Bech-Hansen et al. 2000), evidently involving DBCs only and not HBC circuitry from the ERG results presented here. Further suggestions were that nyctalopin might be important for establishing or maintaining functional connections between the rod photoreceptors and the postsynaptic bipolar cells (Pusch et al. 2000), and again it remains unknown how only DBCs and not HBC synapses are targeted. Hence, it is worthwhile to develop further evidence that human CSNB1 involves only the ON-pathway activity. In the context of the present study implicating DBCs and not HBC, it will be useful to learn whether nyctalopin specifically localizes to the soma or the synapse.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: Correspondence: Paul Sieving, MD, Ph.D., National Eye Institute, Bldg 31 –Room 6A03,31 Center Drive, MSC 2510, Bethesda, MD 20892-2110, Phone: 301-496-2234, Fax: 301-496-9970, Email: paulsieving{at}nei.nih.gov
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Alexander KR, Barnes CS, and Fishman GA. High-frequency attenuation of the cone ERG and ON-response deficits in X-linked retinoschisis. Invest Ophthalmol Vis Sci 42: 2094–2101, 2001a.
Alexander KR, Barnes CS, and Fishman GA. ON-pathway dysfunction and timing properties of the flicker ERG in carriers of X-linked retinitis pigmentosa. Invest Ophthalmol Vis Sci 44: 4017–4025, 2003.
Alexander KR, Barnes CS, Fishman GA, and Milam AH. Nature of the cone ON-pathway dysfunction in melanoma-associated retinopathy. Invest Ophthalmol Vis Sci 43: 1189–1197, 2002.
Alexander KR, Fishman GA, Barnes CS, and Grover S. On-response deficit in the electroretinogram of the cone system in X- linked retinoschisis. Invest Ophthalmol Vis Sci 42: 453–459, 2001b.
Alexander KR, Fishman GA, and Grover S. Temporal frequency deficits in the electroretinogram of the cone system in X-linked retinoschisis. Vision Res 40: 2861–2868, 2000.[CrossRef][ISI][Medline]
Allen LE, Zito I, Bradshaw K, Patel RJ, Bird AC, Fitzke F, Yates JR, Trump D, Hardcastle AJ, and Moore AT. Genotype-phenotype correlation in British families with X linked congenital stationary night blindness. Br J Ophthalmol 87: 1413–1420, 2003.
Antonarakis SE. Recommendations for a nomenclature system for human gene mutations. Nomenclature Working Group. Hum Mutat 11: 1–3, 1998.[CrossRef][ISI][Medline]
Awatramani G, Wang J, and Slaughter MM. Amacrine and ganglion cell contributions to the electroretinogram in amphibian retina. Vis Neurosci 18: 147–156, 2001.[CrossRef][ISI][Medline]
Ball SL, Pardue MT, McCall MA, Gregg RG, and Peachey NS. Immunohistochemical analysis of the outer plexiform layer in the nob mouse shows no abnormalities. Vis Neurosci 20: 267–272, 2003.[ISI][Medline]
Baylor DA, Nunn BJ, and Schnapf JL. The photocurrent, noise and spectral sensitivity of rods of the monkey Macaca fascicularis. J Physiol 357: 575–607, 1984.
Bech-Hansen NT, Naylor MJ, Maybaum TA, Sparkes RL, Koop B, Birch DG, Bergen AA, Prinsen CF, Polomeno RC, Gal A, Drack AV, Musarella MA, Jacobson SG, Young RS, and Weleber RG. Mutations in NYX, encoding the leucine-rich proteoglycan nyctalopin, cause X-linked complete congenital stationary night blindness. Nat Genet 26: 319–323, 2000.[CrossRef][ISI][Medline]
Bergen AA, ten Brink JB, Riemslag F, Schuurman EJ, Meire F, Tijmes N, and de Jong PT. Conclusive evidence for a distinct congenital stationary night blindness locus in Xp21.1. J Med Genet 33: 869–872, 1996.[Abstract]
Berson EL. Retinitis pigmentosa. The Friedenwald Lecture. Invest Ophthalmol Vis Sci 34: 1659–1676, 1993.
