- Email: [email protected]
Retinal mechanisms determine the subadditive response to polychromatic light by the human circadian system
Neuroscience Letters 438 (2008) 242–245
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Retinal mechanisms determine the subadditive response to polychromatic light by the human circadian system Mariana G. Figueiro ∗ , Andrew Bierman, Mark S. Rea Lighting Research Center, Rensselaer Polytechnic Institute, 21 Union Street, Troy, NY 12180, USA
a r t i c l e
i n f o
Article history: Received 21 December 2007 Received in revised form 29 February 2008 Accepted 12 April 2008 Keywords: Melatonin suppression Photoreceptors Additivity Melanopsin Circadian phototransduction Monochromatic and polychromatic light
a b s t r a c t Light is the major synchronizer of circadian rhythms to the 24-h solar day. The intrinsically photosensitive retinal ganglion cells (ipRGCs) play a central role in circadian regulation but cones also provide, albeit indirectly, input to these cells. In humans, spectrally opponent blue versus yellow (b–y) bipolar cells lying distal to the ganglion cell layer were hypothesized to provide direct input to the ipRGCs and therefore, the circadian system should exhibit subadditivity to some types of polychromatic light. Ten subjects participated in a within-subjects 3-night protocol. Three experimental conditions were employed that provided the same total irradiance at both eyes: (1) one unit of blue light (max = 450 nm, 0.077 W/m2 ) to the left eye plus one unit of green light (max = 525 nm, 0.211 W/m2 ) to the right eye, (2) one unit of blue light to the right eye plus one unit of green light to the left eye, and (3) 1/2 unit of blue light plus 1/2 unit of green light to both eyes. The first two conditions did not differ significantly in melatonin suppression while the third condition had significantly less melatonin suppression than conditions 1 and 2. Furthermore, the magnitudes of suppression were well predicted by a previously published model of circadian phototransduction incorporating spectral opponency. As was previously demonstrated, these results show that the human circadian system exhibits a subadditive response to certain polychromatic light spectra. This study demonstrates for the first time that subadditivity is due to spectrally opponent (color) retinal neurons. © 2008 Elsevier Ireland Ltd. All rights reserved.
Light incident on the retina is the primary exogenous stimulus to the circadian system. A variety of empirical studies have demonstrated that, unlike the photopic luminous efficiency function based upon the visual system and used in measurements of light, the spectral sensitivity of the circadian system peaks at short wavelengths (circa max = 450 nm) [2,3,9–12,14]. Further, the absolute sensitivity of the circadian system is much lower than that for the visual system, so a complete transfer function relating optical radiation in the retina to circadian response needs to be developed to accurately predict the effects of light on human health and wellbeing. Studies of retinal neuroanatomy and neurophysiology have shown that a novel photoreceptor, the intrinsically photosensitive retinal ganglion cell (ipRGC), is central to circadian phototransduction [1]. Rods and cones also provide input to the circadian system through the ipRGCs although, clearly, they must do so indirectly through vertical and horizontal connections in the outer plexiform layer of the retina. ON bipolar cells and amacrine cells appear to pro-
∗ Corresponding author. Tel.: +1 518 687 7100; fax: +1 518 687 7120. E-mail address: fi[email protected] (M.G. Figueiro). URL: http://www.lrc.rpi.edu (M.G. Figueiro). 0304-3940/$ – see front matter © 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2008.04.055
vide the strongest connections to the ipRGCs and, therefore, form the photic links between classical photoreceptors and the circadian system [16]. Figueiro et al. [4,5] were the first to conclude that human circadian phototransduction exhibits spectral opponency. Spectral opponency is formed by the three cone types (L, M and S) converging on two classes of bipolar cells that in turn provide separate input to two classes of spectrally opponent ganglion cells [7]. One class of bipolar cells forms the red versus green (r–g) channel with opposing input from L and M cones. The other class forms the blue versus yellow (b–y) channel from S cones opposed to the combined input from the L and M cones. The r–g spectral opponent channel can signal either “red” or “green” and the b–y spectral opponent channel can signal either “blue” or “yellow.” Together with the achromatic luminance channel, these two spectral opponent channels determine a person’s perception of brightness. Brightness perception is inherently subadditive in its response to light due to input from the spectral opponent channels. For example, adding a “green” light to a “red” light will decrease the response of the r–g system, desaturating the vividness of the red light and reducing the apparent brightness of the red light even though more light has been added to it. Like brightness perception, the circadian system can also exhibit a subadditive response to light that presumably like brightness per-
M.G. Figueiro et al. / Neuroscience Letters 438 (2008) 242–245
ception, is the result of bipolar spectral opponent mechanisms in the outer plexiform layer of the retina. Rea et al. [9] modeled the neural connections for human circadian phototransduction using spectrally opponent, S-ON cone bipolars (b–y), lying distal to the ipRGCs. Consistent with conventional neurophysiology, OFF bipolar cells do not provide significant input to the depolarizing ipRGCs [1,7]; certainly, the empirical published data [2,14] cannot be easily modeled with input from a hyperpolarizing OFF bipolar to the depolarizing ipRGCs. Since the model was published, several experiments have been conducted whereby the results are consistent with model predictions of subadditivity. The presumed retinal site for subadditivity has not, however, been formally tested. The present study was specifically designed to determine if the subadditive response to light exhibited by the human circadian phototransduction is formed in the retina. Since spectral opponency leading to the subadditive response to light by the visual system is initiated in the retina, distal to the retinal ganglion cell (RGC) layer (including the ipRGCs), subadditivity should only be demonstrated by combining spectrally opponent lights in one or both eyes, but subadditivity should not be exhibited if these same two lights were presented separately to each eye. On the other hand, if circadian phototransduction is additive, any fraction (f) of a unit irradiance of one light, combined with the complementary fraction (1 − f) of a unit irradiance of a spectrally opponent light should result in the same melatonin suppression regardless if it was mixed in both eyes or given separately to each eye. In the present experiment, unit amounts of a blue light from light emitting diodes (LEDs) (max = 450 nm, 0.077 W/m2 ) and of a green light (max = 525 nm, 0.211 W/m2 ) were determined from the model to produce the same melatonin suppression after a 45-min exposure [9]. Three experimental conditions were employed: (1) one unit of blue light to the left eye plus one unit of green light to the right eye, (2) one unit of blue light to the right eye plus one unit of green light to the left eye, and (3) 1/2 unit of blue light plus 1/2 unit of green light to both eyes. Since the total irradiance to both eyes was the same for conditions 1 and 2, melatonin suppression was hypothesized to be the same; from the model, the predicted percent change in melatonin levels (ratio of melatonin levels before and after light exposure) was 52% [9]. If, as the model predicts, the spectral opponent S-ON cone bipolar (b–y) cells in the retina affect circadian response to light, then condition 3 was hypothesized to produce only a 37% change in melatonin because this is the only condition where fractional amounts of the two light sources are presented to both eyes. If subadditivity is not formed in the retina then all three conditions should produce the same melatonin suppression. Special goggles were each outfitted with four blue (max = 450 nm, 0.077 W/m2 ) and four green (max = 525 nm, 0.211 W/m2 ) Luxeon Rebel (Philips LumiLeds, San Jose, CA) light emitting diodes. Two LEDs of each color were edge-mounted to two separate white acrylic light-mixing diffusers, one for each eye, which extended from the goggle frame to cover most of the subject’s visual field. Fig. 1 shows the spectral power distributions of the lights used in the three experimental conditions. Corneal irradiances (Fig. 1) needed to deliver prescribed doses of light for each condition were calculated [9] taking into account pupil dilation. The calculations were based on the mathematical model developed by Rea et al. [9]. Ten subjects, six males (20–55 years) and four females (18–51 years) participated in the 3-night protocol. Subjects underwent an ophthalmologic exam before the experiment because their pupils were dilated for the study. Two groups of subjects were selected for the study based upon their responses to the Munich Chronotype Questionnaire. Those who habitually fell asleep on weekdays between 22:00 and 00:00 were assigned to group 1, and those who
243
Fig. 1. Corneal illuminances and irradiances for one eye. Total irradiances on both eyes were the same for all three experimental conditions.
fell habitually asleep on weekdays between 23:00 and 01:00 were assigned to group 2. This segregation helped to ensure that the experimental conditions were applied after melatonin onset in all subjects. Subjects were asked to avoid napping and to refrain from caffeine intake after 10:00 on the day of the experiment. The experiment was approved by Rensselaer Polytechnic Institute’s Institute Review Board (IRB). The study was conducted from 23:00 to 01:30 for group 1 and from 00:00 to 02:30 for group during the months of October and November 2007. Unit amounts of blue light emitting diodes (max = 450 nm, 0.077 W/m2 ) and green LEDs (max = 525 nm, 0.211 W/m2 ) light were calculated to produce the same percent change in melatonin levels after a 45-min exposure [9]. Three experimental conditions were employed: (1) one unit of blue light to the left eye plus one unit of green light to the right eye, (2) one unit of blue light to the right eye plus one unit of green light to the left eye, and (3) 1/2 unit of blue light plus 1/2 unit of green light to both eyes. Every subject saw all three experimental conditions, each condition on a different Friday night. Experimental conditions were counterbalanced across subjects. Upon arrival, subjects in both groups signed the consent form and had an in-dwelling angio-catheter inserted into one of their arms by the registered nurse. A single drop of 0.5% cyclopentolate HCl was placed in each of the subjects’ eyes. The first blood draw (3 samples, 4 ml each) was obtained after 45 min in dim room light (less than 0.5 lx at the cornea from red LED lights max = 630 nm). In group 1, the first blood draw was at 00:25. In group 2, the first blood draw was at 01:25. Each subject was then fitted with a light goggle and exposed to a prescribed lighting condition for 45 min after which another blood draw was obtained. Blood samples were spun at approximately 1000 g for 15 min and the plasma samples were frozen at −20 ◦ C; Neuroscience Inc., Osceola, WI performed the melatonin radio-immunoassays. The detection limit of the assays was 1.5 pg/ml. The intra-assay coefficients of variability (CVs) were 12.1% at 16.5 pg/ml, 5.7% at 68.7 pg/ml, and 9.8% at 162.7 pg/ml. The
244
M.G. Figueiro et al. / Neuroscience Letters 438 (2008) 242–245
Fig. 2. Predicted percentage changes in nocturnal melatonin following 45 min of continuous exposure to the three experimental conditions [9], together with the corresponding observed median (square) and mean (triangle) values, including the standard errors of the mean. Photographs of the light-delivering goggles are presented for the three experimental conditions.
inter-assay CVs were 13.2% at 17.3 pg/ml, 8.4% at 69 pg/ml, and 9.2% at 164.7 pg/ml. Data from two subjects on one night each (one datum point for condition 2 and one datum point for condition 3) were excluded from analysis due to failure to follow the experimental protocols on those nights.1 Percent change in melatonin levels was calculated using the ratio of melatonin levels before and after the light exposure. The results of the study, together with the model predictions, are presented in Fig. 2. Mean melatonin suppression ± standard error of the mean (S.E.M.) after exposure to conditions 1, 2 and 3, was 48 ± 9%, 46 ± 10% and 34 ± 7%, respectively. The median values for lighting conditions 1, 2 and 3 were 55%, 62% and 39%, respectively. Both mean and median percent changes in melatonin levels were close to predicted values from the model. A one-way ANOVA showed a significant change in melatonin levels for the three lighting conditions (p < 0.05). Post hoc paired t-test showed no significant change in melatonin levels between conditions 1 and 2 (p > 0.05), but a statistically significant reduced change in melatonin levels in condition 3 compared to those in the other two conditions (p < 0.05). The results presented here provide direct support for a subadditive response to light by the circadian system. These data also demonstrate that the source of subadditivity is in the retina, consistent with retinal neuroanatomy and neurophysiology. More generally, this study is also consistent with predictions generated by Rea et al. [9]. The model generates values of circadian stimulus (CS) for any spectral power distribution (i.e., for any light source at any irradiance). CS is characterized by a high absolute threshold to optical radiation with a peak spectral response at short wavelengths. The model also accounts for participation of ipRCGs as well as rods and cones in circadian phototransduction via neural connections, including spectral opponency formed in the outer plexiform layer of the retina [7]. These results demonstrated that the model can be used to make a priori predictions of circadian response to light of any spectral power distribution. In this context, it is important to note that not every published study of the effect of light on circadian response has shown evi-
1 One subject fell asleep during one session and the other attended the experiment immediately after a party.
dence for spectral opponency [12]. In fact, model predictions show that polychromatic light sources may or may not exhibit subadditivity depending upon the spectral power distribution of the light source. According to the model, contribution of the S-ON cone bipolar in circadian phototransduction can only be revealed as a decrease in the magnitude of the response relative to the magnitude of the response expected from an additive system. The ipRGCs have been shown to only depolarize in response to light [1] and they primarily respond to a depolarizing (“blue” channel response) input from the S-ON cone bipolar [16]. Therefore, if the S-ON bipolar generates a hyperpolarizing (“yellow” channel response) signal, the ipRGC cannot respond directly to this signal. Depending upon their spectral power distributions, polychromatic light sources can produce either a net depolarizing response by the S-ON cone bipolar that adds to the ipRGC response to light or a net hyperpolarizing response by the S-ON bipolar that adds no direct input to the intrinsic light response of the ipRGCs. Thus, nearly any polychromatic light source of sufficient irradiance will produce a circadian response, but only light sources dominated by short-wavelength energy will be able to add to the intrinsic response by the ipRGCs. Although the data presented here are from a limited set of subjects, the results and statistical analyses support the conclusion that the human circadian phototransduction exhibits a subadditive response to some polychromatic light sources due to spectral opponency that is formed in the retina. These results challenge the existing additive measurement techniques for characterizing light and demonstrate that special methods are needed to accurately measure circadian light. Before a complete understanding of the impact of light on human health and well-being can be achieved then, better quantitative measures of circadian light stimulus need to be developed. Without a quantitative understanding of the circadian light stimulus it will be difficult or impossible to make significant progress in unraveling the role that light-induced circadian disruption has on diseases such as breast cancer [13], cardiovascular disease [17], diabetes [6,15] and sleep disorders [8]. Acknowledgements This study was supported by the James D. Watson Young Investigator Award from The New York State Foundation for Science, Technology and Innovation (NYSTAR) to MGF. The authors would like to acknowledge J. Bullough, D. Guyon, T. Klein, L. Lyman, B. Morgan, C. Munson, B. Plitnick, R. Qi, N. Skinner of the Lighting Research Center, Rensselaer Polytechnic Institute, and J. Cunningham and K. Kubarek of The Sage Colleges for their work in this project. References [1] D.M. Berson, F.A. Dunn, M. Takao, Phototransduction by retinal ganglion cells that set the circadian clock, Science 295 (2002) 1070–1073. [2] G.C. Brainard, J.P. Hanifin, J.M. Greeson, B. Byrne, G. Glickman, E. Gerner, M.D. Rollag, Action spectrum for melatonin regulation in humans: evidence for a novel circadian photoreceptor, J. Neurosci. 21 (2001) 6405–6412. [3] C. Cajochen, C. Jud, M. Munch, S. Kobialka, A. Wirz-Justice, U. Albrecht, Evening exposure to blue light stimulates the expression of the clock gene PER2 in humans, Eur. J. Neurosci. 23 (2006) 1082–1086. [4] M.G. Figueiro, J.D. Bullough, A. Bierman, M.S. Rea, Demonstration of additivity failure in human circadian phototransduction, Neuro. Endocrinol. Lett. 26 (2005) 493–498. [5] M.G. Figueiro, J.D. Bullough, R.H. Parsons, M.S. Rea, Preliminary evidence for spectral opponency in the suppression of melatonin by light in humans, Neuroreport 15 (2004) 313–316. [6] K.L. Knutson, A.M. Ryden, B.A. Mander, E. Van Cauter, Role of sleep duration and quality in the risk and severity of type 2 diabetes mellitus, Arch. Intern. Med. 166 (2006) 1768–1774. [7] H. Kolb, How the retina works, Am. Sci. 91 (2003) 28–35.
M.G. Figueiro et al. / Neuroscience Letters 438 (2008) 242–245 [8] A.D. Laposky, J. Bass, A. Kohsaka, F.W. Turek, Sleep and circadian rhythms: key components in the regulation of energy metabolism, FEBS Lett. 582 (2007) 142–151. [9] M.S. Rea, M.G. Figueiro, J.D. Bullough, A. Bierman, A model of phototransduction by the human circadian system, Brain Res. Brain Res. Rev. 50 (2005) 213–228. [10] V.L. Revell, J. Arendt, L.F. Fogg, D.J. Skene, Alerting effects of light are sensitive to very short wavelengths, Neurosci. Lett. 399 (2006) 96–100. [11] V.L. Revell, J. Arendt, M. Terman, D.J. Skene, Short-wavelength sensitivity of the human circadian system to phase-advancing light, J. Biol. Rhythm 20 (2005) 270–272. [12] V.L. Revell, D.J. Skene, Light-induced melatonin suppression in humans with polychromatic and monochromatic light, Chronobiol. Int. 24 (2007) 1125–1137.
245
[13] R.G. Stevens, M.S. Rea, Light in the built environment: potential role of circadian disruption in endocrine disruption and breast cancer, Cancer Cause Control 12 (2001) 279–287. [14] K. Thapan, J. Arendt, D.J. Skene, An action spectrum for melatonin suppression: evidence for a novel non-rod, non-cone photoreceptor system in humans, J. Physiol. 535 (2001) 261–267. [15] E. Van Cauter, J.D. Blackman, D. Roland, J.P. Spire, S. Refetoff, K.S. Polonsky, Modulation of glucose regulation and insulin secretion by circadian rhythmicity and sleep, J. Clin. Invest. 88 (1991) 934–942. [16] K.Y. Wong, F.A. Dunn, D.M. Graham, D.M. Berson, Synaptic influences on rat ganglion-cell photoreceptors, J. Physiol. 582 (2007) 279–296. [17] M.E. Young, M.S. Bray, Potential role for peripheral circadian clock dyssynchrony in the pathogenesis of cardiovascular dysfunction, Sleep Med. 8 (2007) 656–667.
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Retinal mechanisms determine the subadditive response to polychromatic light by the human circadian system Mariana G. Figueiro ∗ , Andrew Bierman, Mark S. Rea Lighting Research Center, Rensselaer Polytechnic Institute, 21 Union Street, Troy, NY 12180, USA
a r t i c l e
i n f o
Article history: Received 21 December 2007 Received in revised form 29 February 2008 Accepted 12 April 2008 Keywords: Melatonin suppression Photoreceptors Additivity Melanopsin Circadian phototransduction Monochromatic and polychromatic light
a b s t r a c t Light is the major synchronizer of circadian rhythms to the 24-h solar day. The intrinsically photosensitive retinal ganglion cells (ipRGCs) play a central role in circadian regulation but cones also provide, albeit indirectly, input to these cells. In humans, spectrally opponent blue versus yellow (b–y) bipolar cells lying distal to the ganglion cell layer were hypothesized to provide direct input to the ipRGCs and therefore, the circadian system should exhibit subadditivity to some types of polychromatic light. Ten subjects participated in a within-subjects 3-night protocol. Three experimental conditions were employed that provided the same total irradiance at both eyes: (1) one unit of blue light (max = 450 nm, 0.077 W/m2 ) to the left eye plus one unit of green light (max = 525 nm, 0.211 W/m2 ) to the right eye, (2) one unit of blue light to the right eye plus one unit of green light to the left eye, and (3) 1/2 unit of blue light plus 1/2 unit of green light to both eyes. The first two conditions did not differ significantly in melatonin suppression while the third condition had significantly less melatonin suppression than conditions 1 and 2. Furthermore, the magnitudes of suppression were well predicted by a previously published model of circadian phototransduction incorporating spectral opponency. As was previously demonstrated, these results show that the human circadian system exhibits a subadditive response to certain polychromatic light spectra. This study demonstrates for the first time that subadditivity is due to spectrally opponent (color) retinal neurons. © 2008 Elsevier Ireland Ltd. All rights reserved.
Light incident on the retina is the primary exogenous stimulus to the circadian system. A variety of empirical studies have demonstrated that, unlike the photopic luminous efficiency function based upon the visual system and used in measurements of light, the spectral sensitivity of the circadian system peaks at short wavelengths (circa max = 450 nm) [2,3,9–12,14]. Further, the absolute sensitivity of the circadian system is much lower than that for the visual system, so a complete transfer function relating optical radiation in the retina to circadian response needs to be developed to accurately predict the effects of light on human health and wellbeing. Studies of retinal neuroanatomy and neurophysiology have shown that a novel photoreceptor, the intrinsically photosensitive retinal ganglion cell (ipRGC), is central to circadian phototransduction [1]. Rods and cones also provide input to the circadian system through the ipRGCs although, clearly, they must do so indirectly through vertical and horizontal connections in the outer plexiform layer of the retina. ON bipolar cells and amacrine cells appear to pro-
∗ Corresponding author. Tel.: +1 518 687 7100; fax: +1 518 687 7120. E-mail address: fi[email protected] (M.G. Figueiro). URL: http://www.lrc.rpi.edu (M.G. Figueiro). 0304-3940/$ – see front matter © 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2008.04.055
vide the strongest connections to the ipRGCs and, therefore, form the photic links between classical photoreceptors and the circadian system [16]. Figueiro et al. [4,5] were the first to conclude that human circadian phototransduction exhibits spectral opponency. Spectral opponency is formed by the three cone types (L, M and S) converging on two classes of bipolar cells that in turn provide separate input to two classes of spectrally opponent ganglion cells [7]. One class of bipolar cells forms the red versus green (r–g) channel with opposing input from L and M cones. The other class forms the blue versus yellow (b–y) channel from S cones opposed to the combined input from the L and M cones. The r–g spectral opponent channel can signal either “red” or “green” and the b–y spectral opponent channel can signal either “blue” or “yellow.” Together with the achromatic luminance channel, these two spectral opponent channels determine a person’s perception of brightness. Brightness perception is inherently subadditive in its response to light due to input from the spectral opponent channels. For example, adding a “green” light to a “red” light will decrease the response of the r–g system, desaturating the vividness of the red light and reducing the apparent brightness of the red light even though more light has been added to it. Like brightness perception, the circadian system can also exhibit a subadditive response to light that presumably like brightness per-
M.G. Figueiro et al. / Neuroscience Letters 438 (2008) 242–245
ception, is the result of bipolar spectral opponent mechanisms in the outer plexiform layer of the retina. Rea et al. [9] modeled the neural connections for human circadian phototransduction using spectrally opponent, S-ON cone bipolars (b–y), lying distal to the ipRGCs. Consistent with conventional neurophysiology, OFF bipolar cells do not provide significant input to the depolarizing ipRGCs [1,7]; certainly, the empirical published data [2,14] cannot be easily modeled with input from a hyperpolarizing OFF bipolar to the depolarizing ipRGCs. Since the model was published, several experiments have been conducted whereby the results are consistent with model predictions of subadditivity. The presumed retinal site for subadditivity has not, however, been formally tested. The present study was specifically designed to determine if the subadditive response to light exhibited by the human circadian phototransduction is formed in the retina. Since spectral opponency leading to the subadditive response to light by the visual system is initiated in the retina, distal to the retinal ganglion cell (RGC) layer (including the ipRGCs), subadditivity should only be demonstrated by combining spectrally opponent lights in one or both eyes, but subadditivity should not be exhibited if these same two lights were presented separately to each eye. On the other hand, if circadian phototransduction is additive, any fraction (f) of a unit irradiance of one light, combined with the complementary fraction (1 − f) of a unit irradiance of a spectrally opponent light should result in the same melatonin suppression regardless if it was mixed in both eyes or given separately to each eye. In the present experiment, unit amounts of a blue light from light emitting diodes (LEDs) (max = 450 nm, 0.077 W/m2 ) and of a green light (max = 525 nm, 0.211 W/m2 ) were determined from the model to produce the same melatonin suppression after a 45-min exposure [9]. Three experimental conditions were employed: (1) one unit of blue light to the left eye plus one unit of green light to the right eye, (2) one unit of blue light to the right eye plus one unit of green light to the left eye, and (3) 1/2 unit of blue light plus 1/2 unit of green light to both eyes. Since the total irradiance to both eyes was the same for conditions 1 and 2, melatonin suppression was hypothesized to be the same; from the model, the predicted percent change in melatonin levels (ratio of melatonin levels before and after light exposure) was 52% [9]. If, as the model predicts, the spectral opponent S-ON cone bipolar (b–y) cells in the retina affect circadian response to light, then condition 3 was hypothesized to produce only a 37% change in melatonin because this is the only condition where fractional amounts of the two light sources are presented to both eyes. If subadditivity is not formed in the retina then all three conditions should produce the same melatonin suppression. Special goggles were each outfitted with four blue (max = 450 nm, 0.077 W/m2 ) and four green (max = 525 nm, 0.211 W/m2 ) Luxeon Rebel (Philips LumiLeds, San Jose, CA) light emitting diodes. Two LEDs of each color were edge-mounted to two separate white acrylic light-mixing diffusers, one for each eye, which extended from the goggle frame to cover most of the subject’s visual field. Fig. 1 shows the spectral power distributions of the lights used in the three experimental conditions. Corneal irradiances (Fig. 1) needed to deliver prescribed doses of light for each condition were calculated [9] taking into account pupil dilation. The calculations were based on the mathematical model developed by Rea et al. [9]. Ten subjects, six males (20–55 years) and four females (18–51 years) participated in the 3-night protocol. Subjects underwent an ophthalmologic exam before the experiment because their pupils were dilated for the study. Two groups of subjects were selected for the study based upon their responses to the Munich Chronotype Questionnaire. Those who habitually fell asleep on weekdays between 22:00 and 00:00 were assigned to group 1, and those who
243
Fig. 1. Corneal illuminances and irradiances for one eye. Total irradiances on both eyes were the same for all three experimental conditions.
fell habitually asleep on weekdays between 23:00 and 01:00 were assigned to group 2. This segregation helped to ensure that the experimental conditions were applied after melatonin onset in all subjects. Subjects were asked to avoid napping and to refrain from caffeine intake after 10:00 on the day of the experiment. The experiment was approved by Rensselaer Polytechnic Institute’s Institute Review Board (IRB). The study was conducted from 23:00 to 01:30 for group 1 and from 00:00 to 02:30 for group during the months of October and November 2007. Unit amounts of blue light emitting diodes (max = 450 nm, 0.077 W/m2 ) and green LEDs (max = 525 nm, 0.211 W/m2 ) light were calculated to produce the same percent change in melatonin levels after a 45-min exposure [9]. Three experimental conditions were employed: (1) one unit of blue light to the left eye plus one unit of green light to the right eye, (2) one unit of blue light to the right eye plus one unit of green light to the left eye, and (3) 1/2 unit of blue light plus 1/2 unit of green light to both eyes. Every subject saw all three experimental conditions, each condition on a different Friday night. Experimental conditions were counterbalanced across subjects. Upon arrival, subjects in both groups signed the consent form and had an in-dwelling angio-catheter inserted into one of their arms by the registered nurse. A single drop of 0.5% cyclopentolate HCl was placed in each of the subjects’ eyes. The first blood draw (3 samples, 4 ml each) was obtained after 45 min in dim room light (less than 0.5 lx at the cornea from red LED lights max = 630 nm). In group 1, the first blood draw was at 00:25. In group 2, the first blood draw was at 01:25. Each subject was then fitted with a light goggle and exposed to a prescribed lighting condition for 45 min after which another blood draw was obtained. Blood samples were spun at approximately 1000 g for 15 min and the plasma samples were frozen at −20 ◦ C; Neuroscience Inc., Osceola, WI performed the melatonin radio-immunoassays. The detection limit of the assays was 1.5 pg/ml. The intra-assay coefficients of variability (CVs) were 12.1% at 16.5 pg/ml, 5.7% at 68.7 pg/ml, and 9.8% at 162.7 pg/ml. The
244
M.G. Figueiro et al. / Neuroscience Letters 438 (2008) 242–245
Fig. 2. Predicted percentage changes in nocturnal melatonin following 45 min of continuous exposure to the three experimental conditions [9], together with the corresponding observed median (square) and mean (triangle) values, including the standard errors of the mean. Photographs of the light-delivering goggles are presented for the three experimental conditions.
inter-assay CVs were 13.2% at 17.3 pg/ml, 8.4% at 69 pg/ml, and 9.2% at 164.7 pg/ml. Data from two subjects on one night each (one datum point for condition 2 and one datum point for condition 3) were excluded from analysis due to failure to follow the experimental protocols on those nights.1 Percent change in melatonin levels was calculated using the ratio of melatonin levels before and after the light exposure. The results of the study, together with the model predictions, are presented in Fig. 2. Mean melatonin suppression ± standard error of the mean (S.E.M.) after exposure to conditions 1, 2 and 3, was 48 ± 9%, 46 ± 10% and 34 ± 7%, respectively. The median values for lighting conditions 1, 2 and 3 were 55%, 62% and 39%, respectively. Both mean and median percent changes in melatonin levels were close to predicted values from the model. A one-way ANOVA showed a significant change in melatonin levels for the three lighting conditions (p < 0.05). Post hoc paired t-test showed no significant change in melatonin levels between conditions 1 and 2 (p > 0.05), but a statistically significant reduced change in melatonin levels in condition 3 compared to those in the other two conditions (p < 0.05). The results presented here provide direct support for a subadditive response to light by the circadian system. These data also demonstrate that the source of subadditivity is in the retina, consistent with retinal neuroanatomy and neurophysiology. More generally, this study is also consistent with predictions generated by Rea et al. [9]. The model generates values of circadian stimulus (CS) for any spectral power distribution (i.e., for any light source at any irradiance). CS is characterized by a high absolute threshold to optical radiation with a peak spectral response at short wavelengths. The model also accounts for participation of ipRCGs as well as rods and cones in circadian phototransduction via neural connections, including spectral opponency formed in the outer plexiform layer of the retina [7]. These results demonstrated that the model can be used to make a priori predictions of circadian response to light of any spectral power distribution. In this context, it is important to note that not every published study of the effect of light on circadian response has shown evi-
1 One subject fell asleep during one session and the other attended the experiment immediately after a party.
dence for spectral opponency [12]. In fact, model predictions show that polychromatic light sources may or may not exhibit subadditivity depending upon the spectral power distribution of the light source. According to the model, contribution of the S-ON cone bipolar in circadian phototransduction can only be revealed as a decrease in the magnitude of the response relative to the magnitude of the response expected from an additive system. The ipRGCs have been shown to only depolarize in response to light [1] and they primarily respond to a depolarizing (“blue” channel response) input from the S-ON cone bipolar [16]. Therefore, if the S-ON bipolar generates a hyperpolarizing (“yellow” channel response) signal, the ipRGC cannot respond directly to this signal. Depending upon their spectral power distributions, polychromatic light sources can produce either a net depolarizing response by the S-ON cone bipolar that adds to the ipRGC response to light or a net hyperpolarizing response by the S-ON bipolar that adds no direct input to the intrinsic light response of the ipRGCs. Thus, nearly any polychromatic light source of sufficient irradiance will produce a circadian response, but only light sources dominated by short-wavelength energy will be able to add to the intrinsic response by the ipRGCs. Although the data presented here are from a limited set of subjects, the results and statistical analyses support the conclusion that the human circadian phototransduction exhibits a subadditive response to some polychromatic light sources due to spectral opponency that is formed in the retina. These results challenge the existing additive measurement techniques for characterizing light and demonstrate that special methods are needed to accurately measure circadian light. Before a complete understanding of the impact of light on human health and well-being can be achieved then, better quantitative measures of circadian light stimulus need to be developed. Without a quantitative understanding of the circadian light stimulus it will be difficult or impossible to make significant progress in unraveling the role that light-induced circadian disruption has on diseases such as breast cancer [13], cardiovascular disease [17], diabetes [6,15] and sleep disorders [8]. Acknowledgements This study was supported by the James D. Watson Young Investigator Award from The New York State Foundation for Science, Technology and Innovation (NYSTAR) to MGF. The authors would like to acknowledge J. Bullough, D. Guyon, T. Klein, L. Lyman, B. Morgan, C. Munson, B. Plitnick, R. Qi, N. Skinner of the Lighting Research Center, Rensselaer Polytechnic Institute, and J. Cunningham and K. Kubarek of The Sage Colleges for their work in this project. References [1] D.M. Berson, F.A. Dunn, M. Takao, Phototransduction by retinal ganglion cells that set the circadian clock, Science 295 (2002) 1070–1073. [2] G.C. Brainard, J.P. Hanifin, J.M. Greeson, B. Byrne, G. Glickman, E. Gerner, M.D. Rollag, Action spectrum for melatonin regulation in humans: evidence for a novel circadian photoreceptor, J. Neurosci. 21 (2001) 6405–6412. [3] C. Cajochen, C. Jud, M. Munch, S. Kobialka, A. Wirz-Justice, U. Albrecht, Evening exposure to blue light stimulates the expression of the clock gene PER2 in humans, Eur. J. Neurosci. 23 (2006) 1082–1086. [4] M.G. Figueiro, J.D. Bullough, A. Bierman, M.S. Rea, Demonstration of additivity failure in human circadian phototransduction, Neuro. Endocrinol. Lett. 26 (2005) 493–498. [5] M.G. Figueiro, J.D. Bullough, R.H. Parsons, M.S. Rea, Preliminary evidence for spectral opponency in the suppression of melatonin by light in humans, Neuroreport 15 (2004) 313–316. [6] K.L. Knutson, A.M. Ryden, B.A. Mander, E. Van Cauter, Role of sleep duration and quality in the risk and severity of type 2 diabetes mellitus, Arch. Intern. Med. 166 (2006) 1768–1774. [7] H. Kolb, How the retina works, Am. Sci. 91 (2003) 28–35.
M.G. Figueiro et al. / Neuroscience Letters 438 (2008) 242–245 [8] A.D. Laposky, J. Bass, A. Kohsaka, F.W. Turek, Sleep and circadian rhythms: key components in the regulation of energy metabolism, FEBS Lett. 582 (2007) 142–151. [9] M.S. Rea, M.G. Figueiro, J.D. Bullough, A. Bierman, A model of phototransduction by the human circadian system, Brain Res. Brain Res. Rev. 50 (2005) 213–228. [10] V.L. Revell, J. Arendt, L.F. Fogg, D.J. Skene, Alerting effects of light are sensitive to very short wavelengths, Neurosci. Lett. 399 (2006) 96–100. [11] V.L. Revell, J. Arendt, M. Terman, D.J. Skene, Short-wavelength sensitivity of the human circadian system to phase-advancing light, J. Biol. Rhythm 20 (2005) 270–272. [12] V.L. Revell, D.J. Skene, Light-induced melatonin suppression in humans with polychromatic and monochromatic light, Chronobiol. Int. 24 (2007) 1125–1137.
245
[13] R.G. Stevens, M.S. Rea, Light in the built environment: potential role of circadian disruption in endocrine disruption and breast cancer, Cancer Cause Control 12 (2001) 279–287. [14] K. Thapan, J. Arendt, D.J. Skene, An action spectrum for melatonin suppression: evidence for a novel non-rod, non-cone photoreceptor system in humans, J. Physiol. 535 (2001) 261–267. [15] E. Van Cauter, J.D. Blackman, D. Roland, J.P. Spire, S. Refetoff, K.S. Polonsky, Modulation of glucose regulation and insulin secretion by circadian rhythmicity and sleep, J. Clin. Invest. 88 (1991) 934–942. [16] K.Y. Wong, F.A. Dunn, D.M. Graham, D.M. Berson, Synaptic influences on rat ganglion-cell photoreceptors, J. Physiol. 582 (2007) 279–296. [17] M.E. Young, M.S. Bray, Potential role for peripheral circadian clock dyssynchrony in the pathogenesis of cardiovascular dysfunction, Sleep Med. 8 (2007) 656–667.