Postnatal development of flicker sensitivity in guinea pigs

OPTOMETRY ORIGINAL PAPER

Postnatal development of flicker sensitivity in guinea pigs James A Armitage MOptom Bang Viet Bui MOptom Riki Gibson BSc BOptom Algis J Vingrys BScOptom PhD Department of Optometry and Vision Sciences, The University of Melbourne

Accepted for publication: 21 September 2001

Background:The retinal response to flickering stimuli (steady state ERG) recruits many retinal elements and is a sensitive indicator of early retinal dysfunction. This study reports the post-natal maturation of the steady state ERG response in guinea pigs. Methods: The steady state ERG response to flickering stimuli (0.6 to 20 Hz) was recorded from dark adapted (more than 12 hrs) English Shorthair guinea pigs ( n = 7 ) using flashes that produced rod and cone dominated responses. Temporal sensitivity functions and critical fusion frequencies (CFF) were derived over a range of ages from postnatal day (PND) 1 to 45. Results: Guinea pig rod and cone temporal sensitivity functions show shape characteristics and CFF similar to humans. Furthermore, the post-natal development of the guinea pig temporal characteristics is also similar to that of humans-they are present at birth and mature rapidly post-natally. The time-course of CFF maturation is similar for rod and cone mediated responses. Conclusions: These data show that the temporal response and its maturation in the guinea pig retina is similar to that in humans. Therefore, we propose that the guinea pig is a particularly usefd animal model to study retinal disease in early childhood. (Clin Exp Optom 2001; 84: 5 : 270-275)

Key words: CFF, development, ERG, flicker, growth, guinea pig

Guinea pigs ( Cauiaporcellus), like humans, are born with their eyes open and a functional visual system.' Given that retinal development of these two species occurs predominantly in ulrro' and as they share a similar rod-to-cone ratio,' various authors have suggested that the guinea pig may be a good model for human development." Indeed, previous developmental work in the guinea pig has demonstrated that humans and guinea pigs show similar maturation of various flash electroretinogram (ERG) components." The flash ERG is a mass measure of lightinduced current fluxes across the retina

a n d includes t h e activity of photoreceptors, neurons, glia arid epithelia.*." However, the flash ERG assays only a small portion of the retinal response and provides little insight in to the temporal capacity of the retina (for review see Weisinger, Vingrys and Sinclair"). The steady state flicker ERG elicited using flickering stimuli can be used to assess the temporal characteristics of the retina. Steady state waveforms are believed to be a summation of the a-wave (photoreceptors), b-wave (rapid O N response) and d-wave (slower OFF response) components of the flash ERG' (Figure l a ) . (:lirurnl nnd Expel inientnl Optornetry 84.5 Septtnilxr 2001

270

Baker and colleagues"' have shown that the main fluxes contributing to the flicker response are located in the inner retina. Moreover, pharmacological inhibition of glutamatergic synaptic transmission" and spiking neurons'2 alters the steady state ERG, thus confirming the presence of postreceptoral inputs to these waveforms. Hence, the flicker ERG can sample the activity of various inner retinal neuronal classes by exploiting their differential temporal and luminance response characteristics. For example, slow flicker rates and dim stimulus intensities will preferentially stimulate n e u r o n s activated by rods,

Development of flicker responses ArmZtap, h i , Gibson and Vingrys

whereas bright fast flicker will favour cone pathways. Similar approaches have been used clinically in the detection and differential diagnosis of retinal disease, including x-linked retinoschisis,"," age-related maculopathy,I5 retinitis pigmentosa,'"," dietary fatty acid depletion," central retinal vein occlusion'"."" and lead toxicity." The flicker ERG has been used to define the post-natal maturation of temporal responses in human infants. In particular, Westall, Panton and Levin" have shown that the amplitude of the human steady state ERG response increases rapidly from birth to five years of age, after which it remains relatively constant. Birch and Anderson'" and Wright and Drasdo" have also reported a similar rapid maturation to adult sensitivity. Given previous reports of similarities between human and guinea pig flash ERG development," we hypothesised that maturation of temporal response properties in the guinea pig may parallel that of humans. Therefore, we investigated the post-natal changes in the guinea pig steady state ERG response.

Figure la. Representative ERG waveform for a long duration stimulus shows the photoreceptoral a-wave, ON- and OFF- responses (called the b- and d-waves)

Figure lb. When the stimulus flickers, a steady state retinal response emerges to stabilise after 10 cycles

60 45

n

30 5

MATERIALS AND METHODS

N

v

i

Animals Pigmented guinea pigs (English Shorthair, n = 7 ) were fed chow ad libitum and housed under lighting of 150 to 300 lux (12 hour cycle, on at 8 am). Although rats reared under similar lighting can show photic retinal damage," guinea pigs under the same conditions appear resistant to light damage.'' Animal housing was maintained at a constant temperature of 22 degrees I:. All experimental procedures were in accordance with the guidelines set out in the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes"" and by o u r institutional ethics committee.

Electroretinogaphy

Figure lc. The linear regression fitted to the descending limb of the peak-to-peak (PTP) amplitude of the steady state response provides an estimate of the critical fusion frequency (CFF).The bright target stimulates the cone system (unfilled symbols) and extrapolation line is fitted to the highest frequency signals (solid line). However, the dim stimulus affects both rod and cone systems (filled symbols), as evidenced by the rod/cone break at approximately 10 to 15 Hz. Here, we fit the extrapolation line to the first limb (broken line).

Figure Id. Maturation of the flicker response is shown by plotting the CFF as a function of age from PND 1 to PND 45 for the same animal

ANIMAL PREPARATION

Prior to measurement, animals were dark adapted overnight (longer than 12 hours), t h e n anaesthetised with a mixture of ketamine (35 mg/kg, Ketamil, Troy

Figure 1. ERG waveforms and their analysis

Clinical and Experimental Optometry 84..5 September 2001

271

20

De\elopment of flicker responws Armrtqy, Buz, Gzbson and Vzngry~

Day 1

Day 7

Day 12

20

v

w 13.3

N

3 10.0

>

5 0

a,

6.7

3 0-

2

LL

3.3 0.7 50 msec Figure 2. Steady state responses elicited by bright stimuli (2 log cd.s.m-2)at various frequencies (numbers on RHS) for the same animal at PND 1 to PND 12. Note the small amplitudes of the waveforms on PND 1 compared to PND 12. Also note that at lower frequencies, there appears to be a double hump that represents the presumed rod (r) and cone (c) contributions to this waveform. Only the initial hump persists at high temporal frequencies (greater than 10 Hz) due to the limited temporal resolution of the rod system.

Laboratories, Smithfield, NSW, Australia) and xylazine (5 mg/kg, Xylazil, Troy Laboratories), delivered by an intramuscular route. Mydriasis (greater than 4 mm) was induced with o n e drop of tropicamide (Mydriacyl 0.5 per cent, Alcon, Frenchs Foi-rest, NSW, Australia) and corneal anaesthesia with one drop ofproxynietacaine (Oplithetic 0.5 per cent, Allergan, Frenchs Forrest, NSW, Australia). Animals were placed in a Faraday cage and ERG signals recorded with a custom Burian-Allen bipolar contact lens rlectrode (Hansen 0 p h t ha1 mic Develo pin en t Laboratory, Iowa, USA) referenced to a stainless steel nredlc electrode, inserted in a neck skinfold. Following electrode placement, a further period of 10 minutes dark adaptation was allowed before recording commenced. All nianipulations "ere carried out under dim red light provided by a red light emitting diode (LED 22 lux @ 10 cm, = 650 t i i n , 23 nm bandwidth at half-height) to preserve dark adaptation.

STIMULUS CALIBRATION

Stiniulus calibration has been described elsewhere." Briefly, our light source was a 150 W (12 V) tungsten-halogen globe coupled to a mechanical shutter/timer with a combined rise and decay of three milliseconds (Uniblitz TS 132, Vincent Associates, Rochester, Nu,USA), The light source was focused on to the face ofa fibre optic cable, which in turn was placed 10 mm from the eye of the animal and focused in to the pupillary plane. This arrangement provided f~illfield (90 degrees) stimulation with an unattenuated stimulus exposure of 2.5 log cd.s.m'. The flash intensity was controlled with calibrated neutral density filters (Schott-Garsco, NSW, Australia). RECORDING CONDITIONS

Full-field, white, flicker ERGSwere elicited over a range of frequencies (0.6, 1.7, 3.3, 6.6, 10,13.3,16.6,20Hz) and at twostiniulus exposures ( - 2 and 2.0 log cd.s.m-?). Stimulus exposures were chosen to isolate (Xnical and Exprriinental 0ptoriietr.y X4.5 Septcrnber 2001

272

rod (-2 log cd.s.ni-2)and cone dominated responses." Signals were amplified and band pass filtered (gain x1000; -3dB at I and 300Hz, 2 cascaded 1st order and a 5th order Bessel filter, Amp 801, Neuroscientific Corp, Farmingdale, NY, USA). For each frequency and light level, the steady state ERG response was measured by averaging the final 60 of 80 flash responses. An inter-stimulus interval of 180 seconds was allowed between each series. Serial ERG measurements were made at a range of post-natal ages (post-natal day (PND) 1 , 5 , 7 , 9 , 12,20,45).

ANALYSIS Rather than performing a discrete Fourier transform, we chose to analyse raw waveforms in terms ofamplitude. We feel that this approach affords a more intuitive interpretation of the data than is possible if analysis is conducted in the frequency domain. The purpose of this modelling was to establish the temporal resolution (critical fusion frequencies, CFF) of the guinea pig eye for rod and cone dominant stimuli. The peak-to-peak amplitude (pV) of the steady state response (Figure l b ) as a function of stimulus frequency gives the temporal sensitivity function (TSF) . There are two methods by which the CFF may be derived. One involves measuring the peakto-peak amplitude at increasing temporal frequencies until this response fails to exceed background noise levels. Alternatively, the CFF can be approximated by modelling a linear rate of change on the descending limb of the temporal sensitivity functiongH(regression line, Figure lc). This latter method is dependent on the spread and number of points used t o derive the slope of this line. DLK to the niechanical limitations of o u r shutter/ timer, we have chosen to iise the latter method for extrapolating the CFF and adopted points at the extreme ends of the response. Although we acknowledge rhat this extrapolation can introduce uncertainty into its estimates, we feel that it will not bias our findings due to the sample size used in the study. Data from each animal were fitted to the three highest temporal frequencies following the peak

Development of flicker responses ArmzLagP, Bui, Gibron. and Vingty

00

PND1 PND7 --t PND12 -+ PND20 0

100

0

1

-

t

'

PND1

T

--+

10

--8-PND7

PND12

-+ PND2O

1

l . . . . : . . . . : . . 0

5

10

15

0

20

5

10

!

15

'

'

'

i

'

20

Frequency (Hz)

Frequency (Hz) Figure 3a. Average temporal sensitivity function under bright stimulus conditions (mean * SEM, n = 7) over several ages. Note, the function shows bandpass characteristicssimilar to the human cone temporal sensitivityfunctions of Hess and S n o ~ d e nNote .~~ also that the function does not change in overall appearance,only increases in amplitude.

FigureSb. Average temporal sensitivity function under dim stimulus conditions (mean * SEM, n = 7). This function shows lowpass characteristicsand is indicative of a rod driven process. Again, the shape of the function does not change with age.

Figure 3. Rod and cone temporal sensitivity functions

sensitivity frequentcy (13, 16, 20 Hz for cones) with the CFF being defined as that point on the abscissa, where the linear regression reaches 10 pV (lower noise limit of the system, grey bar Figure lc). For dim stimuli, a rod-cone break is apparent at about 10 Hz. Therefore, to yield the rod component, we fit to the first three data points following the peak sensitivity (rod component 1.7, 3.3, 6.6 Hz ). Figure Id shows the change in the CFF obtained after such extraction as a function of development. RESULTS Although the guinea pig is a precocial species, it displays a highly immature ERG steady state response at birth (Figure 2, left panel), which increases with age (PND 7, Figure 2, centre panel) and becomes well developed by PND 12 (Figure 2 , right panel). A comparison of' the flicker responses (stimulus intensity 2 log cd.s.m-')

from 1 , 7and 12 day old guinea pigs shows increased amplitudes and faster peak time for all temporal modulations in the older animals. It also shows prominent rod and cone mediated contributions to the waveform. At bright stimulus intensities (2.0 log cd.s.m-') , the temporal sensitivity function (TSF) is band-pass with a peak flicker sensitivity of 7 to 9 Hz. Overall, the shape of the bright TSF does not change appreciably with age (Figure 3a). Rather, there is an upward shift in maximal sensitivity that matures by PND 12. We propose that the light levels and band-pass characteristic of this response is consistent with a cone dominated process.'!' For the dim stimulus exposure (-2 log cd.s.m-2),the TSF has a low-pass profile, with a sensitivity peak between 0.6 and 1.7 Hz (Figure Sb), consistent with a rod dominated response.'9 At bright (Figure 3a, 2.0 1ogcd.s.m-') and dim (Figure 3b, -2 log cd.s.m-') stimulus exposures, the temporal response function Clinical and E x p r r ~ m r n t a Optoinctry l 84 5 September 2001

273

increases with amplitude as a function of age. Response amplitudes increase from PND 1 to PND 12 and then remain relatively constant (PND 20 to 45; data not shown for PND 45). These data have been sunimarised in terms of the critical flicker frequency (CFF) as described previously and shown in Figure 4 as a function of age. Overall, the maturation of processes that determine the CFF under bright (cone dominant, unfilled circles) and dim (rod dominant, filled circles) conditions follow s i m i 1a r d e ve 1 o p men t al ti me c o 11r s e s (Figure 4). There is a rapid increase in CFF (around 40 per cent) from PND 1 to PND 12, after which the CFF attains adult values (Figure 4) and remains invariant up to PND 45. DISCUSSION The flicker electroretinogram (steady state ERG) is a useful method to probe the retinal response of small animals like the

Development of flicker responses Arrnitnp, Uuz, Gibson and I/in.pys

0-L

'

:

0

10

20

30

40

50

Age (days) Figure 4. Development of the CFF. The CFF of both the rod (mean i: SEM, filled symbols) and cone (mean i SEM, unfilled symbols) systems develop with a similar time course. Dotted lines show the average CFF at maturity (defined as average of PND 12 to PND 45).

signalling from rods and cones to second order neurons and we can interpret our findings as indicating a common maturity in rod and cone pathways, as has been reported in humans." Other species, for example cats, demonstrate differential maturation of cone and rod function and may not be good models for human retinal development."4 Previous work and the present findings argue that the guinea pig is an excellent model for human receptoral" a n d post receptoral' as well as temporal tlevelopment. These studies justify the use of the guinea pig for developmental research, particularly for those conditions that compromise developmental maturation, such as chronic placental insufficiencv","' and altered peri-natal dietary intake.",'" ACKNOWLEDGEMENTS The authors wish to thank Associate Pro-

fessor George Smith for assisting with stimulus calibration. This work was supported by a University of Melbourne Research Development Grant (AJV).

guinea pig. The aim of this study was to compare human and guinea pig temporal responses and to assess the suitability of the guinea pig as a niodel for the development of temporal responses in humans. We acknowledge that o u r method of extrapolation for (:FF determination may be prone to some imprecision. However, results show a high level of congruency and any errors appear to be systematic in natiire. Nonetheless, we find that the adult guinea pig temporal sensitivity function, like in humans,:"' has low-pass and bandpass profiles at dim and bright light levels, respectively. Moreover, the maximum teinporal resolution (CFF) of the guinea pig retina improves as a firnction of stimulus intensity (dim: 21 Hz, bright: 51 Hz) as has been demonstrated in humans:"' O u r findings show a pattern of agerelated changes in temporal resolution similar to that demonstrated in human " , I 2 The temporal resolution of the guinea pig reaches maturation by

PND 12 and remains relatively constant into adolescence ( P N D 45). Similarly, hunian infants demonstrate increasing sensitivity over the first five years of' life" and retain peak values into adulthood. Thus, the guinea pig provides a useful model of human CFF development at these early ages. However, flicker sensitivity in aged individuals (older than 40 years) shows a gradual decline"'^".:" that can bc attributed to a decrease in effective retinal illuminance." M7e cannot comment on the possibility of an old-age decline in the guinea pig, as this study followed animals to only relatively young ages (adulthood). The post-natal maturation of rod and cone mediated responses shares a similar time course to that in the guinea pig. Anatomical studies suggest that the expression of light capturing niolecules (opsins) occurs at the same time in both receptor subclasses" even though cones may differentiate earlier.":' This finding may reflect the simultaneous developmental onset of

REFERENCES 1. Huang J, Wysr JP, Spira AM'. Ontogencsis o f the elcctrorrtinogram in a precocial niammal, the guinea pig ( (:civin porrrllus) . Comp Biochrm P'Ijsiol,4 1990; 95: 149-155. 2. Spira AW. In ufrro development and maturation of the retina of a non-primate mammal: a light and electron microscopic study of the guinea pig. AnaLt b,'mbryol ( B ~ r l 1975; ) 146: 279-300. 3. Rees S, Bainbridge A. The structural and neurochemical development of the fetal guinea pig retina antl optic nerve in experimental growth rctardation. / n / J l h i\fwrrtroa/ i 1992; 10: 93.108. 4. vin HofMM', Usarrii E. The ERG in the newborn guinea pig. Proceedings ( ~ tlic f 6th International Symposium on (:linical EK(:. Efiirt, FRG, 1967. 5. U u i RV, Vingrys 41. 1)rvelopnient 01' p ~ h t recepiofiil function i n pigmented and albino guinea pigs. l'zs " V ~ ~ ~ . ~ i 2i 0n 0press. 1; ti. Bui BV, Vingrys .AJ. D e v e l o p m e n t o i receptoral responses i n pigmented antl albino guinea-pigs ( C : / w i c t p o i r r l l i i ) , Ooc Ophthnlmol 1999; 99: 15 I -I 7 0 . 7 . Bui BV, Sinclair 41, Vingrys 41. Electroretinograms of albino antl pigmented guinea-pigs ( C c ~ v i c lporcrl1u.c). Aust .\TZ./ OphltlnlmoE 1998; 26 Snppl 1: s9x-s100. 8 . Granit R. (:ompotwrits o f t h c retinal action potential in mamrnals ;tncl their relations

Development of flicker responses Armitage, Bui, Gibson and Vingrys

to the discharge in the optic nerve.JPhjsio1 1933; 77: 207-238. 9. Weisinger HS, Vingrys AJ, Sinclair AJ. Electrodiagnostic methods in vision Part 2. Origins of the flash ERG. Clin Expw Optom 1996; 79: 97-105. 10. Baker CL Jr, Hess RR, Olsen BT, Zrenner E. Current source density analysis of linear and non-linear components o f t h e primate electroretinogram. JPhysiol1988; 407: 155176. 11. Bush RA, Sieving PA. Inner retinal contributions to the primate photopic fast flicker electroretinogram.JOpt Soc A m A 1996; 13: 557-565. 12. Hood DC, Frishman LJ, Viswanathan S , Robson JG. Ahmed J. Evidence for a ganglion cell contribution to the primate electroretinogram (ERG):effects ofTTX on the multifocal ERG in macaque. Vis Nrurosci 1999; 16: 41 1-466. 13. Alexander KR, Barnes CS, Fishman GA. Origin of deficits in the flicker electroretinogram of the c o n e system in X-linked retinoschisis as derived from response nonlinearities.JOpt Soc A m A 2001; 18: 747754. 14. Alexander KR, Fishman GA, Grover S.Temporal fi-equencydeficits in the electroretintr gram o f the c o n e system in X-linked retinoschisis. Vzsion t@s 2000; 40: 2861-2868. 15. Falsini B, Fadda A, Iarossi G, Piccardi M, Canu D, Minnella A et a]. Retinal sensitivity to flickcr modulation: reduced by early agerelated niaculopathy. Inarst Ophthalmol Vis Sci 2000; 41: 1498-1506. 16. Massof RW,Johnson MA, Sunness JS, Perry C, Finkelstein D. Flicker electroretinogram in retinitis pigmentosa. Doc Ophthnlmol1986; 62: 231-245. 17. Tylcr CW, Ernst U', Lyness AL. Photopic flicker sensitivity losses in simplex and multiplex retinitis pigmentosa. fnvpst Ophlhalmol Vis Sci 1984; 25: 1035-1042. 18. Weisinger HS, Vingrys AJ, Bui BV, Sinclair AJ. Effects of dietary n-3 fatty acid deficiency and repletion in the guinea pig retina. fnvest Ophthalmol VZ5 Sri 1999; 40: 327-338. 19. Ia-sson J , Bauer B, Andreasson S. The 30Hz flicker cone ERG for monitoring the early course of central retinal vein occlusion. Actri Ophthnlmol Scnnd 2000; 78: 187190. 20. Severns ML, J o h n s o n MA. Predicting outcome in central retinal vein occlusion using the flicker electroretinogram. Arch Ophtholmol1993; 111: 1123-1130. 21. Fox DA, Farber DB. Rods are selectively altered by lead. I: Electrophysiology and biochemistry. ExpEyeRPs 1988; 46: 59761 I . 22. Westall CA, Panton CM, Levin AV. Time courses for maturation of electroretinogram responses from infancy to adulthood. Doc Ophthalmol1998;96: 355-379.

23. Birch DG, Anderson JL. Standardized fullfield electroretinography. Normal values and their variation with age. Arch Ophthalmol 1992; 110: 1571-1576. 24. Wright CE, Drasdo N. The influence of age on the spatial and temporal contrast sensitivity function. Doc Ophthalmol1985; 59: 385395. 25. Penn JS,Naash MI, Anderson RE. Effect of light history on retinal antioxidants and light damage susceptibility in the rat. Exp &S 1987; 44: 779-788. 26. National Health and Medical Research Council. Australian code of practice for the care and use of animals for scientific purposes. Canberra: Australian Government Publishing Service, 1990. 27. Nixon PJ. ERG isolation of rod and cone responses in the rat and guinea-pig. BSc (Hons) Thesis, The University of Melbourne, 2000. 28. De I.ange H. Relationship between critical flicker-frequency and a set of low-frequency characteristics of the eye. ,I Opt Soc Am A 1954; 44: 380-389. 29. Hess KF,Snowden RJ. Temporal properties of human visual filters: number, shapes and spatial covariation. Vision Res 1992; 32: 4759. 90. Kelly DH. \'isual responses to timedependent stimuli. I. Amplitude sensitivity measurements.JOpt SorAmA 1961; 51: 422-429. 31. Wolf E, Gaeta AM, Geer SE. Critical flicker frequencies in flicker perimetry. Arch Ophthnlniol1968; 80: 347-352. 32. Tyler CW. Two processes control variations in flicker sensitivity over the life span.] Opt Soc A m A 1989; 6: 481-490. 33. Cepko CL, Austin CP,Yang X, Alexiades M, Ezzeddine D. Cell fate determination in the vertebrate retina. Pror Nut Acnd .Sci LISA 1996; 9Y: 589-595. 34. Jacobson SG, Ikeda H, Ruddock K. Conemediated retinal function in cats during development. Dor Ophthnlmol1987; 65: 7-14. 35. Mallard EC, Rehn A, Rees S, Tolcos M, Copolov D. Ventriculomegaly and reduced hippocampal volume following intrauterine growth-restriction: implications for the aetiology of schizophrenia. Schzzophr Rec 1999; 40: 11-21, 36. Tolcos M, Rees S. Chronic placental insufficiency in the fetal guinea pig affects neurochemica1 and neurogliaI development but not neuronal numbers in the brainstem: a new method for combined stereology and immunohistochemistry.] Comp Neuroll997; 379: 99-1 12. 37. Weisinger HS, Armitage JA, Sinclair AJ, Vingrys AJ, Burns PL, Weisinger RS. Perinatal omega-3 fatty acid deficiency affects blood pressure later in life. Nat Med 2001; 7: 258-259.

Cliiiical d i i d Exprriinerital Oprometry 84.5 September 2001

275

Author's address: Associate Professor Algis J Vingrys Department of Optometry and Vision Sciences T h e University of Melbourne 374 Cardigan St Carlton VIC 3053 AUSTRALIA