Early Inner Retinal Dysfunction in Streptozotocin-Induced Diabetic Rats

Early Inner Retinal Dysfunction inStreptozotocin-Induced Diabetic Rats

Kenichi Kohzaki,1 Algis J. Vingrys,2 and Bang V. Bui2

PURPOSE. Diabetes is known to alter retinal function, as mea-sured with the electroretinogram (ERG), which shows a pro-pensity toward inner retinal oscillatory potential (OPs) abnor-malities. However, the effect that diabetes has on otherganglion cell–related responses is not known. This study was asystematic evaluation of streptozotocin (STZ) diabetes–relatedERG changes in rats for the first 11 weeks after diabetogenesis.

METHODS. Thirty Sprague-Dawley rats were randomly assignedto treated (50 mg/kg STZ (n � 16) and control groups (1 mL/kgcitrate buffer, n � 14) at 6 weeks of age. Two control animalsand four STZ animals were excluded because of blood glucosecriteria or systemic complications. Diabetic animals were givendaily SC injections of 1 to 2 units of long-acting insulin. ERGswere measured at 4, 8, and 11 weeks after treatment. Thea-wave was used as an index of outer retinal function, whereasthe b-wave, OPs, and the scotopic threshold response (STR)were used as indices of inner retinal function.

RESULTS. Photoreceptoral (a-wave) and bipolar cell (b-wave)responses were not significantly reduced by STZ treatment.OPs were significantly reduced by 8 weeks (�25% � 7%, P �0.05). The most severely affected component was the ganglioncell–dominated positive STR, which was significantly de-creased from the first time point (�51% � 11% at 4 weeks, P �0.05), but the negative component was unaffected over the11-week period.

CONCLUSIONS. The ganglion cell dominated pSTR showed largelosses in STZ treated rats. (Invest Ophthalmol Vis Sci. 2008;49:3595–3604) DOI:10.1167/iovs.08-1679

Diabetes can damage neurons,1 glia,2,3 and the vasculartissues within the retina.4,5 Evidence of neuronal alter-

ations includes the presence of apoptosis6,7 in the inner retinallayers and photoreceptors in diabetic animals.7,8 This diabetes-related cell loss leads to a reduction in retinal thickness, whichhas been demonstrated with histopathology,1,6,7 optical coher-ence tomography,9 and scanning laser polarimetry,10 in bothanimals and humans with diabetes. The histologic changes tothe inner retina seem to be more severe.1,6

Clinical evidence for neuronal damage comes from studiesshowing that diabetic patients can have reduced visual acu-ity,11 visual field sensitivity,12 contrast sensitivity,13 color vi-sion,14 and flicker sensitivity.12 Studies in which objectiveelectrophysiological tests, such as the electroretinogram(ERG),15 the pattern ERG (PERG),16 the multifocal ERG,17 and

the visual evoked potential,16 were used have shown abnor-malities before evidence of vascular change in diabetic eyes.

In animals and humans with diabetes, the most commonERG finding is for a reduced amplitude and prolonged implicittime in the oscillatory potentials (OPs) and in some cases analtered photoreceptor response.15 OPs are small-amplitude,high-frequency wavelets, found on the rising slope of theb-wave and thought to involve amacrine cell activity.18 Theearly OP abnormality suggests that these retinal interneuronsare more susceptible to diabetes, especially as the photorecep-tor changes have been shown to be related to the omega-3fatty-acid changes secondary to the diabetes-induced lipidanomaly and not the diabetic hyperglycemic state (Yee P et al.IOVS 2004;45:ARVO E-Abstract 4151). As anatomic studies alsofind apoptosis in ganglion cells early in diabetes,1,19 it is likelythat ERG components reflecting ganglion cell integrity will beaffected along with the OPs.

One ERG component that reflects ganglion cell activity isthe scotopic threshold response (STR). Recent work in rats hasshown that the positive (p)STR reflects ganglion cell activity,whereas the negative (n)STR reflects both ganglion cell andamacrine cell activity.20 Conflicting evidence for STR loss hasbeen reported in humans and animals with diabetes. Aylward21

found in humans that the reduction in STR amplitude andincreased latency correlate with the stage of diabetic retinop-athy. On the other hand, Kaneko et al.22,23 report no STRdeficits in diabetic rats or humans. However, as both investi-gators measured only the nSTR, this may confound contribu-tions from cells other than the ganglion cell.20 Moreover, thenature of the receptoral and postreceptoral inputs to the gan-glion cells have not been well defined in the same diabeticcases. In this study, we considered the nature of ERG changesearly in diabetes with an emphasis on inner retinal changesmanifest in the OPs and STR.

METHODS

Animals

Procedures adhered to the ARVO Statement for the Use of Animals inOphthalmic and Vision Research. Thirty, male Sprague-Dawley ratswere housed in an air-conditioned environment (21°C) with diurnallight cycling (50 lux, 8 AM–8 PM). Food and water were provided adlibitum. At 6 weeks, the rats were randomly assigned to receive a tailvein injection of either 50 mg/kg streptozotocin (n � 16, STZ; MPBiomedicals, Solon, OH) dissolved in trisodium citrate buffer (1 mL/kgof 0.01 M, pH 4.5; Sigma-Aldrich, Castle Hill, NSW, Australia) or bufferalone (n � 14).

Blood glucose levels were measured at 1, 3, 9, and 12 weeks afterSTZ injection (Ascensia Esprit2 and Glucodisc; Bayer HealthCare,Pymble, NSW, Australia), with levels �15 mmol/L indicative of diabe-tes. One week after injection, animals with STZ-induced diabetes (STZanimals) had daily SC injections of insulin (1 to 2 units Protaphane;Novo Nordisk Pharmaceuticals, Baulkham Hills, NSW, Australia) tosustain body weight and general condition and to better mimic thehuman condition. All animals received at least 1 unit of long-actinginsulin, those showing poor grooming and increased urine output asindicated by the condition of the bedding, received 2 units. ThreeSTZ-injected animals were excluded, as blood glucose levels were notsustained above this criterion (�15 mmol/L). Systemic complications

From the 1Department of Ophthalmology, The Jikei University,Minato-ku, Tokyo, Japan; and the 2Department of Optometry andVision Sciences, University of Melbourne, Parkville, Victoria, Australia.

Supported by National Health and Medical Research CouncilGrants 400127 (BVB) and 350224 (AJV).

Submitted for publication January 2, 2008; revised March 11 and31, 2008; accepted May 28, 2008.

Disclosure: K. Kohzaki, None; A.J. Vingrys, None; B.V. Bui, NoneThe publication costs of this article were defrayed in part by page

charge payment. This article must therefore be marked “advertise-ment” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Corresponding author: Algis J. Vingrys, Department of Optometryand Vision Sciences, University of Melbourne, Parkville, 3010, Victoria,Australia; [email protected].

Investigative Ophthalmology & Visual Science, August 2008, Vol. 49, No. 8Copyright © Association for Research in Vision and Ophthalmology 3595

arose in two control and one STZ rat, and these data were excluded.Thus, group data represent the average of 12 control and 12 STZanimals.

Electroretinographic Procedures

ERGs were recorded at 4, 8, and 11 weeks after diabetogenesis usinga Ganzfeld bowl (Photometric Solutions International, Huntingdale,VIC, Australia) containing 20 white LEDs (5-watt, Luxeon; PhilipsLumileds Lighting Co., San Jose, CA). Luminous energy was calibratedusing an integrating photometer (IL1700) with a filter (Z-CIE; Interna-tional Light Technologies Inc., Newburyport, MA) and is specified inscotopic units. LED voltage and duration (1–4 ms) were varied to yieldluminous energies ranging from �6.08 to 1.92 log scot cd � s � m�2.LED temporal profile and conversion from scotopic cd � s � m�2 tophotoisomerizations/rod are given elsewhere.24

At the dimmest energy, 20 responses were averaged, with fewerbeing collected at brighter energies. The interstimulus interval was 2seconds for the dimmest flashes and increased to 120 seconds forbright energies. Responses were filtered 0.1 to 1000 Hz, amplified�1000 (P511; Grass Technologies, West Warwick, RI) and digitized at4 kHz (PowerLab, ADInstruments, Bella Vista, NSW, Australia) over a640-ms epoch. Signals were collected (Scope software, ver. 3.7.6;ADInstruments) for post hoc analysis (Excel; Microsoft, Redmond,WA). Electrodes were silver chloride, one placed on the center of thecornea and the other looped about the equator of the same eye, bothreferenced to a stainless-steel needle (Grass Technologies) inserted inthe tail.

Responses were collected simultaneously from both eyes afterovernight dark-adaptation (�12 hours) and after anesthesia, with anintramuscular injection of a mixture of ketamine hydrochloride (60mg/kg; Troy Laboratories, Smithfield, NSW, Australia) and xylazine (5mg/kg; Ilium Xylazil-100; Troy Laboratories). Both corneas were anes-thetized with 0.5% proxymetacaine hydrochloride (Alcaine; Alcon Lab-oratories, Frenchs Forest, NSW, Australia) and lubricated (Celluvisc;Allergan, Irvine, CA). The pupils were dilated (�4 mm) with 0.5%tropicamide (Mydriacyl; Alcon Laboratories), and body temperaturewas maintained at 37 � 0.5°C, with a water heating pad.24 All proce-dures and electrode placement were performed under dim red illumi-nation (�max � 650 nm).

Photoreceptor Response (P3)

We modeled the leading edge of the a-wave with a delayed Gaussian25

as given by equation 1:

P3�i,t� � RmP3 � 1 � exp� � i � S � �t � td�2� for t � td (1)

This equation describes the P3 response as a function of luminousenergy, i (log cd � s � m�2), and time, t (s). The response is scaled by themaximum saturated amplitude (RmP3, �V). The delay, td (s) of thephototransduction cascade and its sensitivity, S (m2 � cd�1 � s�3),modify this response. Although the td is mainly determined by non-physiological factors, work by our group has shown that this parame-ter can be altered by diabetes,26 and so we floated td. Parameteroptimization was achieved over an ensemble of the three highestenergies (1.32–1.92 log cd � m�2), using the solver module of aspreadsheet (Excel; Microsoft) by minimizing the sum-of-squares (SS)merit function. Past work has shown that the cone contribution to thea-wave at these luminous energies is less than 8%27 and thus negligiblefor our purpose.

Bipolar Cell Response (P2)

The P2 component underlying the b-wave reflects inner retinal func-tion, particularly that of ON-bipolar cells.28 Its was isolated by digitalsubtraction of the P3 model from the raw data to yield the P2–OPcomplex,24 which was then low-pass filtered (�3 dB, 50 Hz) to exposethe P2. P2 amplitude and implicit time were measured from baseline topeak of this waveform and the relationship between P2 amplitude andluminous energy was described by a Naka-Rushton function29:

V�i� � Vmax �i n

i n � K n (2)

The P2 amplitude (microvolts) as a function of luminous energy i (logcd � s � m�2) is described by its saturated response, Vmax (microvolts),a semisaturation constant K (log cd � s � m�2), and a slope n. The slopereturns to 1 for a single underlying cellular generator30 and �1 in thecase of multiple generators.

Inner Retinal Response (OPs and STR)

OPs were extracted from the P2–OP complex with a fifth-order But-terworth band-pass filter (�3 dB; 50–250 Hz). In Sprague-Dawley ratsOP3 is the largest oscillation and the wavelet analyzed for intensities ator brighter than �4.2 log cd � s � m�2. The smooth transition in OPpeak time across intensity (see Figs. 6G–I) is consistent with theisolation of OP3.31 However, OP3 is difficult to identify at dimmerintensities, as the waveform becomes more complex (see Fig. 2C,bottom two traces) with signals that are close to noise levels. As such,we cannot be confident that, at intensities of less than �4.2 log cd � s �

m�2, we have returned OP3.To decrease variability, we averaged OP amplitude over the five

brightest energies (�0.72 log cd � s � m�2) for each animal (Fig. 6,shaded regions). This value was then used to calculate the relative OPchange between treated and control animals (Fig. 7).

The amplitudes of the pSTR and nSTR were returned at fixed timesof 120 and 220 ms (A120 and A220) after stimulus onset. The implicittimes of the pSTR and nSTR were taken from the peak and trough ofthe STR waveform. STR amplitude and implicit time were averagedacross four energies (�6.08 to �5.27 log cd � s � m�2), to increasesignal to noise. These intensities contain minimal intrusion from com-ponents other than the STR.

Statistical Analysis

Group data are specified as the mean � SEM. Normality was deter-mined with a Kolmogorov-Smirnov test, and variance homogeneitywas tested by using a variance ratio. Statistical trends across intensitywere determined using repeated-measures ANOVA (Prism, ver. 4.00;GraphPad Software Inc., San Diego, CA) with Bonferroni post hoc com-parison between groups. For P3 and P2 analysis age (4, 8, and 11 weeks)was nested in treatment (control versus STZ). For other analysis both ageand intensity were nested within treatment. We used an unpaired t-test toevaluate differences in systemic parameters between the control and STZgroups. An � of 0.05 was applied for all statistical purposes.

RESULTS

Figure 1A shows that our STZ-treated animals expressed thetypical diabetic milieu. Although the body weight of the STZgroup increased by 31% (261 � 6 to 341 � 9 g) over the12-week period, control rats showed significantly greaterweight gain (49%; 290 � 5 to 433 � 11 g; F1,3 � 8.37, P �0.001). Post hoc analysis shows that body weights divergedsignificantly after 4 weeks (P � 0.01, Fig. 1A). Figure 1B showsthat STZ animals returned significantly higher blood glucose atall time points (STZ: 25.4 � 0.8 mmol/L versus control 9.7 �1.1 mM, P � 0.001). The elevation in control blood glucose atthe final time point (12.8 � 0.6 mmol/L) occurred because itwas collected with anesthesia needed for tissue harvest andknown to elevate blood glucose.32

Figure 2 shows representative ERG waveforms, at selectedintensities, for a control (thin traces) and a diabetic rat (boldtraces), 11 weeks after treatment. The dimmest three wave-forms show the STR response, where the positive lobe of theSTR was smaller in STZ rats than in control animals (Fig. 2A).For brighter energies, the b-wave was reduced in the STZanimal, whereas the a-wave appeared unaffected (Fig. 2B). TheSTZ animal returned lower OP amplitudes at brighter intensi-ties (� 0.72 log cd � s � m�2), as shown in Figure 2C.

3596 Kohzaki et al. IOVS, August 2008, Vol. 49, No. 8

Photoreceptor Response (P3)

We found a paradoxical increase in RmP3 at 8 weeks in theSTZ group (P � 0.01) which was confirmed by the signifi-cant treatment � time interaction (F1,2 � 4.36, P � 0.02,Fig. 3A). However, as the STZ amplitude returned to controllevels at 11 weeks (control, 536 � 20 �V; STZ, 554 � 36 �V)we interpret this finding as a chance observation. We do notfind the reduced RmP3 reported by others (�16% to �27%),an issue that will be considered later. The decline in RmP3

with age found in the control group is consistent with the

age-related change to RmP3 reported in rats over similar ageranges.33

Phototransduction sensitivity (S) also significantly declinedin both groups at 8 and 11 weeks (P � 0.001). However, thetreatment � time interaction and treatment main effect werenot significant (F1,2 � 0.55, P � 0.58; F1,22 � 2.78, P � 0.11;Fig. 3B) indicating that diabetes does not affect sensitivity ininsulin-treated diabetic rats, as reported by others.26,34,35 Thelatency of the phototransduction cascade (td), was not signifi-cantly different between control and diabetic groups (control,4.54 � 0.06 ms; STZ, 4.55 � 0.07 ms; F1,22 � 2.78, P � 0.11).

FIGURE 1. Average (�SEM) bodyweight and blood glucose for control(n � 12) and STZ animals (n � 12).(A) Body weight was significantly de-creased after 4 weeks in STZ com-pared with control animals. (B)Blood glucose was significantly in-creased at 3, 9, and 12 weeks afterSTZ treatment compared with thecontrol (each n � 5, 2, and 12).Dashed line: normal limit of bloodglucose. Statistically significant: **P �0.01, ***P � 0.001

FIGURE 2. The effect of STZ treat-ment on the full-field ERG waveformsat 11 weeks after STZ injection for arepresentative control (thin traces)and STZ-treated rat (bold traces). (A)Response series to dim luminous en-ergies (�6.08 to �3.09 log cd � s �m�2). (B) Response series to brighterluminous energies (�1.22 to 1.92 logcd � s � m�2). (C) OPs isolated fromflash energies in (B). Luminous ener-gies (log cd � s � m�2) are given to theleft of the waveforms. Scale bars areshown to the top of each group ofwaveforms to which they apply.

IOVS, August 2008, Vol. 49, No. 8 Ganglion Cell Dysfunction in STZ Diabetic Rats 3597

Bipolar Cell Response (P2)

Figure 3 shows that the maximum amplitude of the P2 (Vmax)decreased with age between 4 and 8 weeks for both controland diabetic groups (F1,22 � 5.56, P � 0.01). However, neitherthe treatment � time interaction (F1,2 � 2.20, P � 0.12) northe treatment effect (F1,22 � 1.73, P � 0.20) were significant.Similarly, the semisaturation constant K decreased in bothgroups between 4 and 8 weeks (significant time effect: F2,69 �7.05, P � 0.002; Fig. 3E) indicating improved P2 sensitivity.However, the treatment � time interaction and treatment maineffects were again not significant (F1,2 � 0.69, P � 0.51: F1,22

� 0.04, P � 0.84). Finally, the slope n for the STZ group wassignificantly increased compared with that in the control (treat-ment effect F1,22 � 17.75, P � 0.001, Fig. 3F). However, thelack of a significant treatment � time interaction (F1,2 � 0.22,P � 0.81) indicated no selective time related changes.

Scotopic Threshold Response

Figure 4 shows the peak amplitude (A–C) and implicit times(G–I) for the pSTR. The shaded area of each panel highlightsthe dim stimulus energies (�6.08 to �5.27 log cd � s � m�2) atwhich the STR dominates, as evidenced by a stable implicittime in control animals over these energies (�115 ms, Figs.4G–I). Note that the implicit time shows a transition to aslower peak at approximately 130 ms, presumed to be the rodb-wave, and which in turn transitions to a much faster peak atbright energies (�85 ms), presumed to be the mixed rod–coneb-wave.

The best fit Naka-Rushton model shows that the STZ group(bold curve) had a steeper slope compared with the controlgroup (thin curve, Fig. 4). Normalizing pSTR amplitudes to theaverage control value confirms a greater amplitude reductionat lower stimulus energies in STZ-treated rats (Figs. 4D–F,indicated by the horizontal bar). This accounts for the steeperslope of the Naka-Rushton function in Figures 4A–C. Not sur-prisingly, a significant treatment � intensity interaction wasfound at all weeks (4 weeks: F1,2 � 5.25, P � 0.001; 8 weeks:F1,2 � 11.46, P � 0.001; 11 weeks: F1,2 � 5.67, P � 0.001).

Figures 4G–I show that STZ affected P2 timing differentlyacross energies with a significant treatment x intensity interac-

tion at all weeks (4 weeks: F1,2 � 2.48, P � 0.001; 8 weeks: F1,2

� 2.16, P � 0.001; 11 weeks: F1,2 � 1.76, P � 0.015). Thisfinding was more evident in the normalized timing change inFigures 4J–L. At intermediate intensities (�3.79 to �0.48 logcd � s � m�2), the STZ group had a slower P2, whereas at highintensities (� 0.41 log cd � s � m�2) the P2 was faster than in thecontrol. It is of note that peak time becomes more variable atthe dimmest intensities (gray area) and with the increasingduration of diabetes. This variation may reflect the selectiveloss of the positive STR component, which in the absence ofother positive components becomes difficult to define.

Figure 5 shows the amplitude (A–C) and the implicit time(G–I) for the negative component of the STR. When the fourlowest intensities were analyzed across all weeks, the STZanimals had significantly larger nSTR amplitudes (F1,69 � 6.74,P � 0.015). This result shows a significant treatment � inten-sity interaction (F1,3 � 3.11, P � 0.034). Post hoc analysesrevealed that the nSTR was larger in the STZ group at 4 and 8weeks only (P � 0.05).

In terms of nSTR implicit time, there was a statisticallysignificant treatment � intensity interaction at 11 weeks (F1,2

� 4.70, P � 0.005), but not at 4 and 8 weeks (4 weeks: F1,2 �0.78, P � 0.51; 8 weeks: F1,2 � 0.59, P � 0.62). However,treatment effects were not significant at 4 and 8 weeks (4weeks: F1,22 � 4.40, P � 0.05; 8 weeks: F1,22 � 4.33, P �0.05).

Oscillatory Potentials (OPs)

Figure 6 shows the amplitude (A–C) and implicit time (G–I) ofthe largest OP (or OP3 for intensities at or above �4.2 log cd �s � m�2), with relative amplitude (D�F) and timing change(J�L) also shown. The shaded area of each panel shows thebrighter stimulus energies (0.72–1.92 log cd � s � m�2), whereOP parameters were averaged to give an overall effect. Figure6 shows two phases in OP amplitude growth: The first growthphase began at �4 log cd � s � m�2 where it returns its slowesttiming. The first plateau was at �2.5 log cd � s � m�2, andthereafter the OPs sped up and showed amplitude saturation atthe highest energies.

FIGURE 3. The effect of STZ treatment on P3 and P2 at 4, 8, and 11 weeks. P3 parameters: (A) saturated photoreceptor amplitude (RmP3), (B)sensitivity (S), (C); delay (td) of phototransduction cascade. P2 parameters: (D) maximum amplitude (Vmax); (E) the semisaturated constant (K);(F) the slope (n) of the Naka-Rushton function. Bars, group mean � SEM. Statistically significant: **P � 0.01, ***P � 0.001.


The effect of STZ on OP amplitudes varies as a function ofintensity, with a decrease at higher light levels (Figs. 6D–F) asreflected in the significant treatment � intensity interaction(F1528 � 4.22, P � 0.025). Overall OP amplitude was reducedin the STZ group (Figs. 6A–C, F1,22 � 4.93, P � 0.037), withsignificant reductions in OP amplitudes at 8 and 11 weeks (P �0.05) but not 4 weeks. At 8 weeks, post hoc tests showed thatSTZ OPs were significantly smaller for energies above �1.5 logcd � s � m�2. Post hoc analyses at other ages were not signifi-cant. Figures 6G–I show OP implicit times for all ages. OP peaktime in STZ animals was slower at intermediate energy levels,but faster than in control animals at high energies. AlthoughSTZ appears to slow OPs most at intermediate intensities, theeffect was not significant.

Relative STZ Effect on ERG Components

Figure 7 compares the effect of STZ on the various ERG com-ponents normalized to the average control. Figure 7A showsthat both P2 (Vmax) and P3 (RmP3) amplitudes were littleaffected by diabetes. Averaged over all ages, only the P2 wassignificantly smaller in STZ rats (average; �6% � 3%: F1,22 �10.28, P � 0.004). Figure 7B shows that whereas the P3 timingwas slower (at 8 weeks), P2 peak time was faster at this age.

Figure 7C shows that the nSTR was paradoxically increased at4 and 8 weeks (overall, 22% � 6%), whereas the pSTR wasreduced at all weeks (overall average, �56% � 7%). In addi-tion, the nSTR was 6% � 2% slower (Fig. 7 D) and wassignificant at 11 weeks. Figure 7E shows that OPs were re-duced (F1,2 � 7.20, P � 0.014) at 8 and 11 weeks. However theoverall average OP reduction of �19% � 5% was less than thepSTR. It is also worth noting that at 4 weeks, the only ERGcomponent significantly reduced by STZ treatment was thepSTR. OP and pSTR timing was unchanged (Figure 7F).

DISCUSSION

Photoreceptoral a-Wave (P3)

We found that 11 weeks of STZ-diabetes returned photorecep-tor amplitude, sensitivity and timing parameters (RmP3, S andtd; Fig. 3), that were not significantly different from the control.However, the absence of an a-wave abnormality cannot beruled out, given that, despite a longitudinal experimental de-sign, we needed an experimental power of 59% to find a 20%a-wave difference between the control and STZ-treated group.Nevertheless, normal a-waves have been reported in STZ-

FIGURE 4. Effect of STZ on the first positive peak. (A–C) Show the relationship between peak amplitude and stimulus energy modeled with aNaka-Rushton function (lines) at 4, 8, and 11 weeks, respectively. In the each panel, lines fit control (thin line) and STZ (bold line) data. (D–F)Data normalized to the control mean. Asterisk and bar: the range over which post hoc analysis was significant. (G–I) show the relationshipbetween peak implicit time and stimulus energy at 4, 8, and 11 weeks, respectively. (J–L) Normalized data as before. Shaded area gives the regionfor STR responses recorded from dimmer stimulus energies (�6.08 to �5.27 log cd � s � m�2). Each symbol and error bar represents the mean �SEM in the control (n � 12) and the STZ (n � 12) groups.


treated rats36–38 and human diabetics.39 On the other hand,studies also report photoreceptor deficits in STZ ani-mals26,34,35,40 and humans.41 Differences in the duration ofdiabetes in the above studies may account for the disagreementas to the degree of a-wave loss. Li et al.40 and Phipps et al.34

find significant a-wave changes 10 to 12 weeks after STZinduction and not before. Thus, it is possible that a-wavedeficits may have become apparent if we had evaluated the STZeffect later.

Age-Related Changes in Control andSTZ-Treated Rats

It is worth noting that RmP3 decreased by approximately 15%in both groups between 4 and 11 weeks (age, 10–17 weeks).This decline may be attributed to ageing of the eye. Fulton andHansen33 showed that the rat a-wave reaches its peak ampli-tude at 5 weeks after birth. Kiyosawa42 showed that Wistar ratsundergo a �17% decline in a-wave amplitude between 5 and

17 weeks of age. Likewise, Hancock and Kraft38 reported a28% reduction in a-wave amplitude from 8 to 20 weeks of age.These changes are similar in magnitude to our age-relatedchanges.

Age related ERG amplitude attenuation may arise from eyegrowth. However, Guggenheim et al.43 show that rat axiallength reaches adult dimensions by 10 weeks. Alternatively,Kiyosawa42 demonstrated thinning of outer and inner nuclearlayers in rats between postnatal weeks 5 and 13. Our controlanimals showed the greatest a-wave change between postnatalweeks 10 (4 weeks after STZ) and 14, and only a small reduc-tion between weeks 14 and 17. This functional patternmatches the anatomic trends,42 thus postnatal refinement ofretinal layers may account for our amplitude reduction. Aninteresting finding is that the age-related a-wave decline oc-curred later at 14 weeks of age (8 weeks after treatment) in STZrats. However, as we have only a single time point at 8 weeks,the existence of this trend needs further investigation.

FIGURE 5. Effect of STZ on the aver-age (� SEM) nSTR amplitude and im-plicit time. Amplitudes were taken ata fixed criterion time of 220 ms afterstimulus flash. Data are shown for 4(A), 8 (B), and 11 (C) weeks afterSTZ treatment. (D–F) Normalizedamplitudes. Implicit times were mea-sured to the trough minimum (G–I).(J–L) Relative implicit times. Theshaded region identifies the nSTR re-sponse recorded from dimmer stim-ulus energies (�6.08 to �5.27 logcd � s � m�2).


ON-Bipolar Cell b-Wave (P2)

Li et al.40 reported a �30% b-wave loss at 2 weeks, whereasHancock and Kraft38 reported a 33% b-wave loss 12 weeks afterSTZ-injection. Parisi et al.44 found the b-wave reduced in dia-betic patients. Although, we found only a small b-wave de-crease (�13%) in STZ rats, our finding is similar to those in aprevious report of normal b-wave amplitudes at 8 weeks,34

followed by a b-wave loss of 15% to 18%, 12 weeks after STZinduction. The lack of change in b-wave implicit time in ourstudy is also consistent with results in another study.34

An interesting finding is that the slope of the b-wave inten-sity response function was significantly increased at all times.This increase in slope is consistent with the finding that am-plitude change is greater at dimmer intensities. This issue isdiscussed in the following sections.

Inner Retinal Oscillatory Potentials (OPs)

A robust outcome in diabetes is that the OPs are either reducedor delayed, as reviewed by Shirao and Kawasaki.15 OP deficitsin diabetes were first observed in the early 1960s.45 Severalgroups have since shown that smaller OP amplitudes are asso-ciated with greater retinopathy in humans.21,44,46 Other stud-ies have found OPs to be affected early in STZ diabetes (�5weeks).34,37,38 However, Li et al.40 reported that OP losses

occurred only at later ages (�20 weeks) and after the b-wavewas reduced.

Kizawa et al.47 found that the OPs were more affected bydiabetes than was the a- or b-wave, in a large group ofhuman diabetics. Similarly, animal studies, have found OPsto be more attenuated26 and affected earlier than outer retinalresponses.34,36 This is consistent with our finding that OPswere significantly reduced at 8 (�25% � 7%) and 11 (�22% �11%) weeks, before any losses in the a- or b-waves.

Scotopic Threshold Response

Parisi et al.44 and Aylward21 have reported that OP amplitudesare better indicators of diabetic retinopathy severity than otherwaveforms arising from the inner retina. They show that thePERG (review by Parisi and Uccioli16), which is the ganglioncell response to a contrast-reversing checkerboard48 is lesssensitive in diabetes than are the OPs. Similarly, Kizawa et al.47

showed in human diabetics that OPs are better indicators ofretinopathy progression than is the photopic negative re-sponse (PhNR), which receives contributions from ganglioncells in humans.49 However, Arden et al.50 reported the PERGas the best indicator of diabetic retinopathy. Consistent withArden et al.,50 we found an apparent pSTR deficit 4 weeks afterSTZ treatment, whereas we did not see significantly reducedOPs until 8 weeks.

FIGURE 6. Effect of STZ on the peak OP response. (A–C) Average (� SEM) OP amplitude; (D–F) normalized response at 4, 8, and 11 weeks,respectively. Asterisk and bar: the range over which post hoc analysis was significant. (G–I) Average (� SEM) implicit time; (J–L) normalizedimplicit time. The shaded region shows OP responses recorded from brighter stimulus energies (0.72–1.92 log cd � s � m�2).


The finding that the pSTR loss was greater than the b-wavereduction, together with no loss in the nSTR suggests thatganglion cell dysfunction does not arise from reduced inputfrom distal elements. This contention receives further supportfrom our finding that the slope of the b-wave intensity re-sponse function was significantly steeper (Fig. 3F). A steeperslope has been detected after removal of proximal retinalcontributions, by using -amino butyric acid application inmouse retina.30

Kaneko23 reported normal nSTR in human diabetics and inrats 2 weeks after STZ administration.22 Although, our findingsare in agreement with those of Kaneko et al.,22 the absence ofan nSTR loss (Fig. 7) is paradoxical, given that the nSTRreceives contributions from both ganglion and amacrinecells.20 It is likely that, in early diabetes, losses in positivecomponents such as the pSTR and the P2 mask any reductionin the corneal negative nSTR. However, nSTR reductions maymanifest later in the course of diabetes. Consistent with thispossibility, Aylward21 has found the nSTR to be a good predic-tor of more severe retinopathy in humans. It is also worthnoting that the paradoxical increase in the nSTR argues that theloss in inner retinal function reflects gains between the ON-bipolar cells and inner retinal pathways, as this would manifestas a reduction in all inner retinal waveforms.

It is also worth noting that the glial cell damage known toexist in diabetes2,3 may influence the STR, as glial cell–medi-ated potassium flux is thought to be involved in pSTR and nSTRgeneration.51 However, our finding for a pSTR loss with alarger nSTR suggests that the deficit is unlikely to be mediatedsolely by glial cell dysfunction, as this would affect both pSTRand nSTR to a similar magnitude.

Mechanisms of Retinal Dysfunction

Our finding that the pSTR demonstrated the greatest deficit isconsistent with the increased apoptosis observed in the gan-glion cell layer at 2 to 4 weeks after STZ treatment in rats,1,52

after 2 months in mice,6 and in human diabetics.19 However,increased apoptosis has also been observed in other retinal cellpopulations—in particular, amacrine cells7 and photorecep-tors.8

Several mechanisms may account for increased ganglioncell dysfunction. First increased activation of the polyol path-way53 and sorbitol accumulation54 can alter retinal function.Sorbitol accumulation has been shown to be higher in theganglion cell layer in alloxan-treated rabbits.54 Moreover, gan-glion cell loss is decreased with treatment to reduce sorbitolaccumulation.53 Second, vascular abnormalities specific to theinner retinal capillary plexus would be more detrimental toproximal (amacrine and ganglion cells) than to distal retinalneurons. Do Carmo et al.5 reported that the inner blood–retinalbarrier in diabetic rats is more compromised than is the outerbarrier in STZ rats. Third, glutamate concentration is increasedin the retina of diabetic animals55 and in the vitreous of hu-mans with proliferative diabetic retinopathy.56 Ganglion cellsare particularly susceptible to a sustained extracellular increasein glutamate, leading to excitotoxicity.57

The exact mechanism leading to greater ganglion cell dys-function in STZ rats must be further investigated. Nevertheless,diabetes induced ganglion cell dysfunction may represent arisk factor for glaucoma (reviewed by Toda and Nakanishi-Toda58). In glaucoma, it is thought that mechanical compres-sion of the laminar cribrosa59 leads to reduced retrograde

FIGURE 7. STZ data (mean � SEM)normalized to the average controlvalues at 4, 8, and 11 weeks. Theamplitude and the timing of RmP3

versus the Vmax is shown in (A) and(B), respectively. (C, D) Amplitudeand timing of the nSTR versus thepSTR, respectively. (E, F) Amplitudeand timing of the OPs versus thepSTR, respectively. In each panel,the shaded area represents the nor-mal range for the filled symbols,whereas the solid lines represent thenormal range for open symbols.Dashed line: normal level. Statisti-cally significant: *P � 0.05, **P �0.01, ***P � 0.001.


transport of neurotrophic factors and ganglion cell loss.60 Ab-normal axonal transport has been reported in diabetic ganglioncell axons.53

In summary, we showed that in the presence of minorphotoreceptor and ON-bipolar cell changes, inner retinal func-tion was significantly reduced in STZ-treated rats. The ganglioncell–dominated pSTR was the most sensitive component toSTZ diabetes, manifesting as early as 4 weeks after STZ treat-ment and before OP loss.

References

1. Barber AJ, Lieth E, Khin SA, Antonetti DA, Buchanan AG, GardnerTW. Neural apoptosis in the retina during experimental and hu-man diabetes: early onset and effect of insulin. J Clin Invest.1998;102:783–791.

2. Barber AJ, Antonetti DA, Gardner TW. Altered expression of retinaloccludin and glial fibrillary acidic protein in experimental diabetes.The Penn State Retina Research Group. Invest Ophthalmol Vis Sci.2000;41:3561–3568.

3. Gaucher D, Chiappore JA, Paques M, et al. Microglial changesoccur without neural cell death in diabetic retinopathy. VisionRes. 2007;47:612–623.

4. Bursell SE, Clermont AC, Kinsley BT, et al. Retinal blood flowchanges in patients with insulin-dependent diabetes mellitus andno diabetic retinopathy. Invest Ophthalmol Vis Sci. 1996;37:886–897.

5. Do carmo A, Ramos P, Reis A, Proenca R, Cunha-vaz JG. Break-down of the inner and outer blood retinal barrier in streptozotocin-induced diabetes. Exp Eye Res. 1998;67:569–575.

6. Martin PM, Roon P, Van Ells TK, Ganapathy V, Smith SB. Death ofretinal neurons in streptozotocin-induced diabetic mice. InvestOphthalmol Vis Sci. 2004;45:3330–3336.

7. Gastinger MJ, Singh RS, Barber AJ. Loss of cholinergic and dopa-minergic amacrine cells in streptozotocin-diabetic rat andIns2Akita-diabetic mouse retinas. Invest Ophthalmol Vis Sci. 2006;47:3143–3150.

8. Park SH, Park JW, Park SJ, et al. Apoptotic death of photoreceptorsin the streptozotocin-induced diabetic rat retina. Diabetologia.2003;46:1260–1268.

9. Sugimoto M, Sasoh M, Ido M, Wakitani Y, Takahashi C, Uji Y.Detection of early diabetic change with optical coherence tomog-raphy in type 2 diabetes mellitus patients without retinopathy.Ophthalmologica. 2005;219:379–385.

10. Takahashi H, Goto T, Shoji T, Tanito M, Park M, Chihara E.Diabetes-associated retinal nerve fiber damage evaluated withscanning laser polarimetry. Am J Ophthalmol. 2006;142:88–94.

11. Lopes de Faria JM, Katsumi O, Cagliero E, Nathan D, Hirose T.Neurovisual abnormalities preceding the retinopathy in patientswith long-term type 1 diabetes mellitus. Graefes Arch Clin ExpOphthalmol. 2001;239:643–648.

12. Stavrou EP, Wood JM. Central visual field changes using flickerperimetry in type 2 diabetes mellitus. Acta Ophthalmol Scand.2005;83:574–580.

13. Dosso AA, Bonvin ER, Morel Y, Golay A, Assal JP, Leuenberger PM.Risk factors associated with contrast sensitivity loss in diabeticpatients. Graefes Arch Clin Exp Ophthalmol. 1996;234:300–305.

14. Hardy KJ, Lipton J, Scase MO, Foster DH, Scarpello JH. Detectionof colour vision abnormalities in uncomplicated type 1 diabeticpatients with angiographically normal retinas. Br J Ophthalmol.1992;76:461–464.

15. Shirao Y, Kawasaki K. Electrical responses from diabetic retina.Prog Retin Eye Res. 1998;17:59–76.

16. Parisi V, Uccioli L. Visual electrophysiological responses in personswith type 1 diabetes. Diabetes Metab Res Rev. 2001;17:12–18.

17. Bearse MA Jr, Adams AJ, Han Y, et al. A multifocal electroretino-gram model predicting the development of diabetic retinopathy.Prog Retin Eye Res. 2006;25:425–448.

18. Wachtmeister L. Oscillatory potentials in the retina: what do theyreveal. Prog Retin Eye Res. 1998;17:485–521.

19. Abu-El-Asrar AM, Dralands L, Missotten L, Al-Jadaan IA, Geboes K.Expression of apoptosis markers in the retinas of human subjectswith diabetes. Invest Ophthalmol Vis Sci. 2004;45:2760–2766.

20. Bui BV, Fortune B. Ganglion cell contributions to the rat full-fieldelectroretinogram. J Physiol. 2004;555:153–173.

21. Aylward GW. The scotopic threshold response in diabetic retinop-athy. Eye. 1989;3:626–637.

22. Kaneko M, Sugawara T, Tazawa Y. Electrical responses from theinner retina of rats with streptozotocin-induced early diabetesmellitus (in Japanese). Nippon Ganka Gakkai Zasshi. 2000;104:775–778.

23. Kaneko M. Scotopic threshold response in noninsulin-dependentdiabetes mellitus patients with good HbA1c level and ophthalmo-scopically intact fundus (in Japanese). Nippon Ganka GakkaiZasshi. 2001;105:463–469.

24. Weymouth AE, Vingrys AJ. Rodent electroretinography: methodsfor extraction and interpretation of rod and cone responses. ProgRetin Eye Res. 2008;27(1):1–44.

25. Hood DC, Birch DG. A quantitative measure of the electricalactivity of human rod photoreceptors using electroretinography.Vis Neurosci. 1990;5:379–387.

26. Bui BV, Armitage JA, Tolcos M, Cooper ME, Vingrys AJ. ACEinhibition salvages the visual loss caused by diabetes. Diabetolo-gia. 2003;46:401–408.

27. Nixon PJ, Bui BV, Armitage JA, Vingrys AJ. The contribution ofcone responses to rat electroretinograms. Clin Exp Ophthalmol.2001;29:193–196.

28. Stockton RA, Slaughter MM. B-wave of the electroretinogram: areflection of ON bipolar cell activity. J Gen Physiol. 1989;93:101–122.

29. Naka KI, Rushton WA. S-potentials from luminosity units in theretina of fish (Cyprinidae). J Physiol. 1966;185:587–599.

30. Saszik SM, Robson JG, Frishman LJ. The scotopic threshold re-sponse of the dark-adapted electroretinogram of the mouse.J Physiol. 2002;543:899–916.

31. Layton CJ, Safa R, Osborne NN. Oscillatory potentials and theb-wave: partial masking and interdependence in dark adaptationand diabetes in the rat. Graefes Arch Clin Exp Ophthalmol. 2007;245:1335–1345.

32. Saha JK, Xia J, Grondin JM, Engle SK, Jakubowski JA. Acute hyper-glycemia induced by ketamine/xylazine anesthesia in rats: mecha-nisms and implications for preclinical models. Exp Biol Med (May-wood). 2005;230:777–784.

33. Fulton AB, Hansen RM. Electroretinogram responses and refractiveerrors in patients with a history of retinopathy prematurity. DocOphthalmol. 1995;91:87–100.

34. Phipps JA, Fletcher EL, Vingrys AJ. Paired-flash identification of rodand cone dysfunction in the diabetic rat. Invest Ophthalmol VisSci. 2004;45:4592–4600.

35. Phipps JA, Yee P, Fletcher EL, Vingrys AJ. Rod photoreceptordysfunction in diabetes: activation, deactivation, and dark adapta-tion. Invest Ophthalmol Vis Sci. 2006;47:3187–3194.

36. Sakai H, Tani Y, Shirasawa E, Shirao Y, Kawasaki K. Developmentof electroretinographic alterations in streptozotocin-induced dia-betes in rats. Ophthalmic Res. 1995;27:57–63.

37. Ramsey DJ, Ripps H, Qian H. An electrophysiological study ofretinal function in the diabetic female rat. Invest Ophthalmol VisSci. 2006;47:5116–5124.

38. Hancock HA, Kraft TW. Oscillatory potential analysis and ERGs ofnormal and diabetic rats. Invest Ophthalmol Vis Sci. 2004;45:1002–1008.

39. Lovasik JV, Kergoat H. Electroretinographic results and ocularvascular perfusion in type 1 diabetes. Invest Ophthalmol Vis Sci.1993;34:1731–1743.

40. Li Q, Zemel E, Miller B, Perlman I. Early retinal damage in exper-imental diabetes: electroretinographical and morphological obser-vations. Exp Eye Res. 2002;74:615–625.

41. Holopigian K, Greenstein VC, Seiple W, Hood DC, Carr RE. Evi-dence for photoreceptor changes in patients with diabetic reti-nopathy. Invest Ophthalmol Vis Sci. 1997;38:2355–2365.

42. Kiyosawa I. Age-related changes in visual function and visual or-gans of rats. Exp Anim. 1996;45:103–114.

43. Guggenheim JA, Creer RC, Qin XJ. Postnatal refractive develop-ment in the Brown Norway rat: limitations of standard refractiveand ocular component dimension measurement techniques. CurrEye Res. 2004;29:369–376.


44. Parisi V, Uccioli L, Monticone G, et al. Electrophysiological assess-ment of visual function in IDDM patients. Electroencephalogr ClinNeurophysiol. 1997;104:171–179.

45. Yonemura D, Aoki T, Tsuzuki K. Electroretinogram in diabeticretinopathy. Arch Ophthalmol. 1962;68:19–24.

46. Bresnick GH, Palta M. Oscillatory potential amplitudes: relation toseverity of diabetic retinopathy. Arch Ophthalmol. 1987;105:929–33.

47. Kizawa J, Machida S, Kobayashi T, Gotoh Y, Kurosaka D. Changesof oscillatory potentials and photopic negative response in pa-tients with early diabetic retinopathy. Jpn J Ophthalmol. 2006;50:367–373.

48. Maffei L, Fiorentini A, Bisti S, Hollander H. Pattern ERG in themonkey after section of the optic nerve. Exp Brain Res. 1985;59:423–425.

49. Viswanathan S, Frishman LJ, Robson JG, Harwerth RS, Smith EL3rd. The photopic negative response of the macaqueelectroretinogram: reduction by experimental glaucoma. InvestOphthalmol Vis Sci. 1999;40:1124–1136.

50. Arden GB, Hamilton AM, Wilson-Holt J, Ryan S, Yudkin JS, Kurtz A.Pattern electroretinograms become abnormal when backgrounddiabetic retinopathy deteriorates to a preproliferative stage: possi-ble use as a screening test. Br J Ophthalmol. 1986;70:330–335.

51. Frishman LJ, Steinberg RH. Light-evoked increases in [K�]o inproximal portion of the dark-adapted cat retina. J Neurophysiol.1989;61:1233–1243.

52. Oshitari T, Roy S. Diabetes: a potential enhancer of retinal injury inrat retinas. Neurosci Lett. 2005;390:25–30.

53. Ino-Ue M, Zhang L, Naka H, Kuriyama H, Yamamoto M. Polyolmetabolism of retrograde axonal transport in diabetic rat largeoptic nerve fiber. Invest Ophthalmol Vis Sci. 2000;41:4055–4058.

54. MacGregor LC, Rosecan LR, Laties AM, Matschinsky FM. Alteredretinal metabolism in diabetes. I. Microanalysis of lipid, glucose,sorbitol, and myo-inositol in the choroid and in the individuallayers of the rabbit retina. J Biol Chem. 1986;261:4046–4051.

55. Lieth E, Barber AJ, Xu B, et al. Glial reactivity and impairedglutamate metabolism in short-term experimental diabetic retinop-athy. Penn State Retina Research Group. Diabetes. 1998;47:815–820.

56. Ambati J, Chalam KV, Chawla DK, et al. Elevated gamma-aminobu-tyric acid, glutamate, and vascular endothelial growth factor levelsin the vitreous of patients with proliferative diabetic retinopathy.Arch Ophthalmol. 1997;115:1161–1166.

57. Kuehn MH, Fingert JH, Kwon YH. Retinal ganglion cell death inglaucoma: mechanisms and neuroprotective strategies. Ophthal-mol Clin North Am. 2005;18:383–395, vi.

58. Toda N, Nakanishi-Toda M. Nitric oxide: ocular blood flow, glau-coma, and diabetic retinopathy. Prog Retin Eye Res. 2007;26:205–238.

59. Bellezza AJ, Hart RT, Burgoyne CF. The optic nerve head as abiomechanical structure: initial finite element modeling. InvestOphthalmol Vis Sci. 2000;41:2991–3000.

60. Quigley HA, Addicks EM. Regional differences in the structure ofthe lamina cribrosa and their relation to glaucomatous optic nervedamage. Arch Ophthalmol. 1981;99:137–143.


Documents

Early Inner Retinal Dysfunction in Streptozotocin-Induced Diabetic Rats