Electrophysiology showed that sensitivity of GGA knockdowns was only slightly diminished relative to appropriate controls out to 22 days post-eclosio light-reared. Presence of ERG on-and off-transients showed that R1-6 synaptic connections to the first order visual neuropil (the lamina ganglionaris) were functional. In transmission electron microscopy, structure of the GGAkd retina and first optic neuropil were strikingly normal with virtually no signs of recrptor degeneration at 8 days post-eclosion light-reared. The only obvious abnormality is that some cells had split rhabdomeres. Following out to 37 days-post-eclosion in the light, receptors were still healthy though there was an accumulation of membrane circles and intraretinular pigment granules. Considering the overall normality in electrophysiology and electron microscopy, the deep pseudopupil was non-existant from newly-emerged dark-reared out to >40 days in the light. Using the optical neutralization technique, oddly, though in disarray, rhabdomere tips were present and similar in appearance from newly-emerged dark-reared out to >40 days in the light. Based on histological observations, Eissenberg et al. claimed that GGAkd flies had age-dependent retinal degeneration; my electrophysiology, ultrastructure and optics contradict this conclusion.

ERGs of GGA knockdowns and controls


GGA knockdown - Female GGAE3 X Male eyGal4+GMRGal4. The appropriate control was driver only. The control for the control was wild-type.

Electrophysiology. Electrographic (ERG) methodology analysis has been standard in this laboratory for decades (Stark, 1973) (McKay et al, 1995). For this work, the stimulus was from a 150 W Xenon Arc in an Opti-Quip housing and 1600 power supply and the readout was via a PowerLab 410 feeding into a Macintosh computer running operating system 9.2. To photograph the deep pseudopupil (DPP) before running the ERG, the cover slip the flies were fixed to was optically fused to a microscope slide using a drop of water, and the DPP was viewed with a 10x air objective. For the ERG, a long wavelength, 585 nm, was chosen for two reasons: (1) potential differences between the effects of eye color pigments of experimental and control flies would be minimized (Stark, 1973); and (2) visualization of on- and off transients would be optimized (Stark & Wasserman, 1972). Data acquisition was expedited by quickly obtaining responses to a defined sequence of seven stimuli of 1.0 s 0.3 to 0.6 log units apart (an intensity response sequence). After obtaining ERGs, the cover slip was again placed on a slide, and rhabdomeres were photographed using oil immersion and a 40x objective.



ERGs of GGA knock-downs were surprisingly consistent with the expectation for "normal" flies. Near threshold, i.e. when the ERG component attributed to receptor depolarization was 0-3 mV, there were on- and off-transients that were fairly large. For higher intensities, the receptor wave was larger, and the transients were diminished. At the highest intensities, the receptor component sometimes settled back rather than just swing up to a steady state. Newly emerged GGA knockdown flies were run as well as flies of various ages out to 22 days, and they all had similar ERG responsivities and sensitivities. ERG on- and off-transients not only imply a high level of R1-6 function but also that the synaptic connections of R1-6 in the first order optic neuropile, the lamina ganglionaris, are functional. GGA knock downs, (top) and driver-only controls (second row) are shown. Newly emerged (left), 6-8 days (second column), 15 days (3rd column) and 22 days (right). For comparison, wild-type newly emerged is at the bottom left. The vertical scales have been adjusted to be equivalent

Newly eclosed, knockdown and driver are about the same sensitivity and about 0.6 log units than wild type. With age, the knockdown stays about the same while the driver only stock increases to wild type, consistent with the increase in rhabdomere size in only the driver-only stock. Assuming the lower sensitivity of the GGA knockdown were based on strictly on a lower amount of rhodopsin, then the inference is that the GGA knockdown has about 0.25 x the normal rhodopsin amount since sensitivity and visual pigment content were shown to be linearly related in the invertebrate (Hamdorf & Schwemer, 1975). Contrast the approximate 0.6 log unit diminution of the GGA knockdown's sensitivity with the over 2 log unit decrease of sensitivity of vitamin A deprived flies (Stark & Zitzmann, 1976).


Hamdorf K, Schwemer J (1975) Photoregeneration and the adaptation process in insect photoreceptors of invertebrates. In Photoreceptor optics, Snyder AW, Menzel R (eds), pp 263-289. Berlin: Springer

McKay RR, Chen DM, Miller K, Kim S, Stark WS, Shortridge RD (1995) Phospholipase C rescues visual defect in norpA mutant of Drosophila melanogaster. J Biol Chem 270: 13271-13276

Stark WS (1973) The effect of eye colour pigments on the action spectrum of Drosophila. J Insect Physiol 19: 999-1006

Stark WS, Wasserman GS (1972) Transient and receptor potentials in the electroretinogram of Drosophila. Vision Res 12: 1771-1775

Stark WS, Zitzmann WG (1976) Isolation of adaptation mechanisms and photopigment spectra by vitamin A deprivation in Drosophila. J Comp Physiol 105: 15-27

work completed Summer 2009

Transmission Electron Microscopy


Transmission electron microscopy (TEM). Control (left) and GGA knockdown (right). Flies were aged eight days in normal room lighting after emergence as adults from the pupa case. Roughly equivalent fields of distal retina (top) and receptor axons projecting to the first optic neuropil ("brain") (bottom) were selected for comparison. There is only one striking abnormality, and that concerns the rhabdomeres, the rhodopsin-containing photoreceptive organelles (dark circles in the retina at the top): The number, shape, size and organization for the GGA knockdown (right) does not match the tidy organization of 7 rhabdomeres per ommatidium (facet in the compound eye) for the control (left).

Considering how bad the retina looks optically, in the deep pseudopupil and with optical neutralization of the cornea (see this text), it is astounding how many untrastructural features of the visual system in eight day post-eclosion adults are close to normal. Retinula cells appear healthy and have normal nuclei and mitochondria (1)(3). They are connected to their neighboring retinula cells with the appropriate belt desmosomes (2)(3). Rhabdomeres have the customary submicrovillar cisternae (3). Within microvilli, there is the usual electron dense rod (3). Also retinula cells have normal intraretinular pigment granules (5) and rhabdomere caps (6). The intraretinular pigment granules are known to migrate toward the rhabdomere in the light, and their position is indicative that the cell was responding to light at the time of fixation (5).

Axons from the retina proceed through the basement membrane to the first optic neuropil, the lamina ganglionaris, where the R1-6 terminals form into the appropriate synaptic glomeruli, optic cartridges (7, control left, GGA kd right, bottom of figure). The reason this was considered to be important was that, in rdgB (which has light-induced retinal degeneration), R1-6 terminals in the optic cartridges showed gross degeneration under the minimal dim red light conditions sufficient for dissection for fixation (Stark & Sapp, 1989), well before the retinula cells in the retina fill with a dense reticulum and lipid droplets (Stark & Carlson, 1982). Close examination reveals the membrane specializations that characterize functional synapses, the T-bars (8).

(figure, knockdown left, control right) It is tempting to speculate that the rich investment of membrane circles in the retinula cell cytoplasm near the rhabdomere is a characteristic feature of the GGA kd; however, the control also shows these structures, and it is not realistic to compare them quantitatively. Such circles were posited as the vehicles to carry rhodopsin and/or membrane to rhabdomeres during vitamin A replacement (Stark et al, 1988); also they were plentiful in ora (outer rhabdomeres absent) where there were no rhabdomeres to receive membrane intended for rhabdomeres (Stark & Sapp, 1987) and after retinoic acid feeding (Lee et al, 1996). Importantly, they fill the retinula cell cytoplasm in Rab11 mutant cells that lack rhodopsin transport to the rhabdomeres (Satoh et al, 2005). Such vesicles also fill the cytoplasm in Drosophila Rip11 (Rab11 interacting protein) mutants (Li et al, 2007).

[The knockdown figure (left) shows a split R7 rhabdomere and several gross abnormalities at the 11 and 2 o'clock positions, presumably in secondary pigment cells.]

This list of healthy features focuses our attention on the most striking abnormalities: the size, orientation and number of rhabdomeres in each ommatidium are irregular (2); also there are gaps between ommatidia (9); some gaps may represent fused or fragmented ommatidia, others may have resulted from damage during tissue preparation for fixation. Upon examination of ommatidia with too many rhabdomeres, the retinula cell count is usually correct; thus retinula cells often have too many rhabdomeres (2) (4). While autophagic bodies, large endosomes, abnormal lysosome-related bodies, or disrupted biosynthetic machinery (Golgi apparatus or rough endoplasmic reticulum) might have been expected, no striking alterations from control were present.

TEM of 37 day ey GMR GGA knockdown

Many of the descriptions for 8 day apply here

Bodies somewhat like MVBs but with double membrane bounded vesicles and pits
Here is shown a tremendous quantity of membrane circles and intraretinular pigment granules in a plane near the basement membrane; see discussion of circles here.
Here we saw one (and only one) retinula cell that appeared to be dead or dying.
Here is shown a peculiar basement membrane, very swolen
Here is a distal section, oblique (nearly longitudinal) showing the rhabdomere split and looking like beads on a string
The optic cartridge is shown here with fairly normal structure and conspicuous T-synapses

TEM of 24 day ey GMR GGA knockdown

Here is the abnormal accumulation of circles and pigment granules
Bodies somewhat like MVBs but with double membrane bounded vesicles and pits (as in 37 day)
Here is an indication of normal synaptic structure


Lee RD, Thomas CF, Marietta RG, Stark WS (1996) Vitamin A, visual pigments and visual receptors in Drosophila. Micros Res Tech 35: 418-430

Li BX, Satoh AK, Ready DF (2007) Myosin V, Rab11, and dRip11 direct apical secretion and cellular morphogenesis in developing Drosophila photoreceptors. J Cell Biol 177: 659-669

Satoh AK, O'Tousa JE, Ozaki K, Ready DF (2005) Rab11 mediates post-Golgi trafficking of rhodopsin to the photosensitive apical membrane of Drosophila photoreceptors. Development 132(7): 1487-1497

Stark WS, Carlson SD (1982) Ultrastructural pathology of the compound eye and optic neuropiles of the retinal degeneration mutant (w rdgBKS222) Drosophila melanogaster. Cell Tissue Res 225: 11-22

Stark WS, Sapp RJ (1987) Ultrastructure of the retina of Drosophila melanogaster: The mutant ora (outer rhabdomeres absent) and its inhibition of degeneration in rdgB (retinal degeneration-B). J Neurogenet 4: 227-240

Stark WS, Sapp RJ (1989) Retinal degeneration and photoreceptor maintenance in Drosophila: rdgB and its interaction with other mutants. In Inherited and Environmentally Induced Retinal Degenerations, LaVail MM, Anderson RE, Hollyfield JG (eds), pp 467-489. New York: Liss

Stark WS, Sapp RJ, Schilly D (1988) Rhabdomere turnover and rhodopsin cycle: maintenance of retinula cells in Drosophila melanogaster. J Neurocytol 17: 499-509

Rhabdomere optics


In landmark papers, Franceschini and Kirschfeld (Franceschini & Kirschfeld, 1971a) (Franceschini & Kirschfeld, 1971b) revolutionized the study of fly photoreceptor optics. Because the index of refraction of water or immersion oil approximate that of the cornea, the focusing function of these facet lenslets is eliminated when the eye is viewed with an immersion objective. Also, rhabdomeres carry light like a fiber optic (light pipe). Hence, this "optical neutralization of the cornea" technique allows viewing of rhabdomere tips when light is transmitted up (antidromically) through the eye. Because of the regularity of rhabdomeres in each ommatidium and the regularity of the arrangement of ommatidia, a low power air objective can be used to image the "deep pseudopupil," a virtual image of the rhabdomere tips, magnified deep (about 150 microns beneath the cornea) and superimposed from about 25 ommatidia (that number depending on the numerical aperture of the microscope objective). Stark & Thomas (2004) is a good overall reference


(Here is a plate made from GGA kd 43 days (left) [F1 from male eyGMRdriver x female GGA.E3], control eyGMR 42 days (middle) [eyGMRdriver;boss/+], GGA control 40 days (right) [GGA.E3/+]
Because the deep pseudopupil and the optical neutralization of the cornea techniques are relatively straightforward, they have long served as a convenient entry point to diagnose the structural integrity of the retina. The deep pseudopupil of GGA knockdowns (driven by ey GMR) was in such disarray that there was no focal plane that showed any rhabdomere clearly. This was invariably the case at every time point assayed ranging from zero days (in the dark or in the light) out to 4 weeks. Focusing up and down, the eye color pigment seemed mottled and uneven.

Here is a selected depth series to demonstrate these points from GGA kd 43 days post-eclosion where 1 = distal, 6 = deeper than the deep pseudopupil

In optical neutralization, rhabdomeres, though abnormal, were present (Figure) and were the same on day zero dark reared and 29 days light reared.

Under optical neutralization of the cornea, rhabdomeres were present at all time points. However, rarely was a single rhabdomere imaged clearly. The number, size, and orientation of rhabdomeres that could be imaged per ommatidium was always variable. In contrast, rdgB (retinal degeneration B, even amid a wasteland of dead R1-6 cells and debris) and ora (outer rhabdomeres absent, lacking R1-6 rhabdomeres) gave clear images of R7 (and clearly lacking R1-6) (Harris et al, 1976).


Why were the GGA knockdown's pseudopupils so disrupted? The diagnoses were straightforward when Harris et al (1976) utilized the pseudopupil techniques developed by (Franceschini & Kirschfeld (1971a; b): R1-6 was clear but R 7 was missing in sevenlses (sev); R7 was clear but R1-6 was missing in outer rhabdomeres absent (ora in what Washburn & O'Tousa, (1989) later showed to be a rhodopsin mutation) and in retinal degeneration (rdgB).

With hindsight, it is obvious that the optical neutralization of the cornea presaged the subsequent histological and ultrastructural findings of variable numbers, sizes, counts and configurations of rhabdomeres in the ommatidia. Since the deep pseudopupil is formed by the magnified, superimposed images of the rhabdomere tips collected from many ommatidia, it is reasonable to conclude that all the factors of rhabdomere disarray preclude a good deep pseudopupil image.


Franceschini N, Kirschfeld K (1971a) Etude optique in vivo des elements photorecepteurs dans l'oeil compose de Drosophila. Kybernetik 8: 1-13

Franceschini N, Kirschfeld K (1971b) Les phenomenes de pseudopupille dans l'oeil compose de Drosophila. Kybernetik 9: 159-182

Harris WA, Stark WS, Walker JA (1976) Genetic dissection of the photoreceptor system in the compound eye of Drosophila melanogaster. J Physiol (Lond) 256: 415-439

Stark WS, Thomas CF (2004) Microscopy of multiple visual receptor types in Drosophila. Mol Vis 10: 943-955

Washburn T, O'Tousa JE (1989) Molecular defects in Drosophila rhodopsin mutants. J Biol Chem(264): 15464-15466

Acknowledgements. The work described above was done Summer 2009-Spring 2010 as part of a collaboration with Joel Eissenberg, Department of Biochemistry and Molecular Biology, Saint Louis University School of Medicine. Dr Eissenberg sponsored my SLU-funded sabbatical Spring 2010. The SLU Presidents Research Fund and Beaumont grant funded some of these studies.The hairpin construct allowing RNAi for GGA was provided to Prof. Eissenberg by Andre C. Dennes, DanielaWaschkau and Regina Pohlmann at the UKM Munster. Electron microscopy was done in collaboration with Dr. Jan Ryerse and Ms Barbara Nagel of SLU's Imaging Core; the grids for the 8 day TEM analysis were from Prof. Eissenberg. Many of the GGAkd flies were isolated and delivered to me by Anne Ilvarsonn.