The Phillips Lab

Evidence for a Light-dependent Magnetic Compass

Our laboratory has used behavioral studies of shoreward magnetic compass orientation by Eastern red-spotted newts Notophthalmus viridescens (Phillips 1986a) to characterize the functional properties of the underlying magnetoreception mechanism. These studies have shown that the newt's magnetic compass is sensitive to the axis, but not the polarity, of the magnetic field ("axial sensitivity"). Like migratory birds, newts use inclination or dip angle, rather than polarity, to distinguish between the two ends of the magnetic axis (Phillips 1986b). More recent experiments have shown that newts tested in the dark fail to exhibit a consistent direction of orientation relative to the magnetic field (Phillips & Borland 1992a). Both axial sensitivity and dependence on the presence of light are consistent with theoretical models that have implicated specialized photoreceptors in magnetoreception (Ritz et al. 2000. Biophys. J. 78: 707-718.). Characterizing the effect of different wavelengths of light on the shoreward response of newts has provided evidence for a wavelength-dependent 90 degree shift in the direction of magnetic compass orientation (Phillips & Borland 1992b). By training newts to a shoreward direction under long-wavelength light, we were able to show that the 90 degree shift was the result of a direct effect of light on the underlying magnetoreception mechanism (Fig 1).

Figure 1. Evidence for a Wavelength-Dependent 90 degree shift in the Directional Response of the Newt's Magnetic Compass (data from Phillips & Borland 1992b). Orientation of newts A) trained and tested under full spectrum light, B) trained under full spectrum light and tested under wavelengths > 500 nm, C) trained under wavelengths > 500 nm and tested under full spectrum light, and D) trained and tested under wavelengths > 500 nm. Data are the magnetic bearings of individual newts tested in one of four magnetic field alignments (magnetic north at north, east, south or west). Bearings are plotted relative to the magnetic direction of shore in the outdoor training tanks (= 360 degrees); data from different training tank alignments combined. Arrows at centers indicate the mean vector bearings with the length of each arrow proportional to the mean vector length ("r"; radius corresponds to r = 1). For each condition, a schematic of a training tank and test arena shows predicted effects of exposure to full spectrum (unshaded) or long-wavelength (yellow shaded) light. Thin arrows indicate the true (mN) or 90 degree rotated (mN') alignments of magnetic north. Wide arrows indicate shoreward magnetic direction newts would learn if they perceive magnetic north to be rotated 90 degrees counterclockwise (ccw) under long-wavelength light.

Because the newt's magnetic compass is insensitive to the polarity of the magnetic field (Phillips 1986a), the pattern of response in a circular array of receptors (or in a single receptor that is used to sample different alignments relative to the magnetic field) is expected to be radially symmetrical. Given such a pattern, we hypothesized that the wavelength-dependence of the newt's magnetic compass response could result from an antagonistic interaction between a short-wavelength mechanism and a less-sensitive, long-wavelength mechanism (Fig 2; and see Phillips & Borland 1992b). Alignments of the magnetic field that produce (e.g.) an increase in response under short-wavelength light would produce a decrease in response under long-wavelength light. This would result in "complimentary bimodal patterns" in which the axes having the highest (or lowest) level of response differ by 90 degrees. [Note that the sign of the response under short-wavelength light (i.e., an increase or decrease) is not important, only that the responses are opposite in sign under short- and long-wavelength light.] If this hypothesis is correct, equal excitation of the two spectral mechanisms at an intermediate wavelength should cause the two complimentary patterns to cancel out and prevent newts from obtaining directional information from the magnetic field. As predicted, magnetic orientation was abolished under intermediate wavelengths around 475 nm (Phillips & Borland 1992b).

Figure 2. Complimentary Bimodal Pattern Hypothesis (Phillips & Borland 1992a). Top diagrams show patterns of response in a hypothetical magnetoreception system consisting of a circular array of axially sensitive receptors (small rectangles). Under short-wavelength light (left upper diagram), receptors in specific alignments relative to either end of the magnetic axis exhibit an increase in response to light (white rectangles labeled "+") relative to receptors in alignments unaffected by the magnetic field (gray rectangles). Arrowheads at the edge of the circular array indicate the axis with the highest level of response. Under an intermediate wavelength of light (center diagram) that activates the two mechanisms more-or-less equally (see lower diagram), the response of the receptors is independent of magnetic field alignment. Under long-wavelength light (right diagram), receptors in alignments affected by the magnetic field exhibit a decrease in response to light due to the antagonistic effect of the long-wavelength mechanism (dark rectangles labeled "-") relative to receptors unaffected by the magnetic field. The axes with the highest level of response differ by 90 degrees under short-wavelength and long-wavelength light. The lower diagram shows the absorption spectra of two hypothetical photoreception mechanisms proposed to mediate the newt's light-dependent magnetic compass. Because the response of newts tested under full spectrum light was indistinguishable from that of newts tested short-wavelength light, the short-wavelength mechanism is assumed to be more sensitive than the long-wavelength mechanism.

Evidence for a wavelength-dependent 90 degree shift has also been obtained in fruit flies Drosophila melanogaster (Phillips & Sayeed 1993) and, more recently, in Bullfrog tadpoles Rana catesbeiana (Freake & Phillips in prep.). The bullfrog findings are interesting because evidence for similar wavelength-dependent magnetic compasses in urodele (salamanders) and anuran (frogs and toads) amphibians, groups thought to have diverged over 350 million years ago, suggest that this type of mechanism is ancestral in amphibians.

Localization of the Light-Dependent Magnetic Compass in Amphibians

Behavioral studies of newts using small spectral filters attached to the top of the head have provided evidence that the newt's magnetic compass is mediated by extraocular photoreceptors located in or near the pineal organ (Fig 3; and see Deutschlander et al. 1999a, Deutschlander et al. 1999b, Phillips et al. 2001). Interestingly, Dodt & Heerd (1966. J. Neurophysiol. 25: 405-429) characterized chromatic units in the pineal complex of frogs (Rana) with spectral properties that closely match those of the light-dependent magnetic compass in newts (Fig 2). Evidence for a wavelength-dependent 90 degree shift in the shoreward magnetic compass orientation of bullfrog tadpoles (Freake and Phillips in prep.) similar to that observed in newts (Fig 1), suggests that the magnetic compasses of these two groups of amphibians are mediated by similar magnetoreception mechanisms. Neurophysiological recordings are now underway in the laboratory to characterize the response of units in the pineal complex of bullfrogs to changes in the alignment of an earth-strength magnetic field.

Figure 3. Evidence for the Involvement of Extraocular Photoreceptors in the Light-Dependent Magnetic Compass Orientation of Newts (from Deutschlander et al. 1999a,b). A-C) Predicted effects of exposure to full spectrum (unshaded) or long-wavelength (yellow shaded) light during training and testing on the shoreward magnetic compass response of newts. The predicted direction of orientation relative to the real (mN) or perceived (mN') direction of magnetic North is shown by double-headed arrows. D-H) Magnetic bearings obtained after 'one day' training from newts tested in one of four horizontal alignments of the magnetic field. Unlike newts used in earlier experiments that were trained for 5-7 days (Fig 1), newts trained for only one day in the outdoor tanks exhibit bimodal magnetic orientation (Deutschlander et al. 2002. Copeia in press). D) Newts trained and tested under full spectrum light. E) Newts trained under full spectrum light and tested under long-wavelength light. F) Newts trained and tested under long-wavelength light. G-H) Newts trained under full spectrum light and tested under long-wavelength light with either clear (G) or long-wavelength transmitting (H) "caps" covering the pineal.

Biophysical Basis of the Light-Dependent Magnetic Compass

A theoretical model proposed by Ritz et al. (2000. Biophys. J. 78: 707-718) implicates a magnetically sensitive radical pair reaction occurring in a specialized photoreceptor as the basis of the light-dependent magnetic compass (Ritz et al. 2002). Alternative mechanisms involving particles of magnetite have also been proposed (Edmonds 1996. Proc. Roy. Soc. Lond. B 263: 295-298; Kirschvink et al., 2001. Cur. Opin. Neurobiol. 11: 462-467). To discriminate between photoreceptor-based (radical pair) and magnetite-based mechanisms, behavioral experiments are currently underway in our laboratory at Virginia Tech to investigate the effects of low level broad band radio frequencies on the magnetic compass orientation of newts and mice. Recent experiments with European robins, carried out in collaboration with the Wiltschko laboratory at the University of Frankfurt, Germany, have shown that magnetic compass orientation in birds is abolished by exposure to low-level radio frequencies in the 1-10 MHz range (Ritz et al. 2004).

"Genetic Dissection" of the Light-Dependent Magnetic Compass

Behavioral studies of Drosophila have provided evidence for a wavelength-dependent 90 degree shift in magnetic orientation that is consistent with the involvement of a light-dependent magnetic compass (Phillips & Sayeed 1993). Experiments are currently underway with mutant strains of Drosophila to investigate the role of the compound eye and ocelli, and other more specialized photoreception mechanisms (e.g., cryptochromes), in mediating the light-dependent magnetic compass. We are also developing an assay of magnetic compass orientation in the C57 BL/6 strain of laboratory mouse that will be used for behavioral genetic analyses. The mouse assay is based on an approach that we have used to demonstrate magnetic compass orientation in the Siberian Hamster Phodopus Sungorus (Deutschlander et al. 2003).

This material is based upon work supported by the National Science Foundation under Grant No. IBN97-24083

Back to Home