282
KENNETH V. THIMANN AND GEORGE M. CURRY
In the action spectrum the peak in the near ultraviolet suggests
participation by riboflavin. Quantitatively, however, it is much smaller
than one would expect if riboflavin alone were the photoreceptor, because the absorption peak in the near ultraviolet of riboflavin, and of
other common flavin derivatives, is little, if at all, lower than that in
the visible. Studies of the spectral sensitivity for the photolysis of IAA
in the presence of riboflavin or concentrated brei of etiolated pea
epicotyls give action spectra closely resembling the absorption spectrum
of riboflavin (98, 121). The curves rise sharply below 410 τημ to a
height at 390 τημ surpassing the 440 τημ maximum in the visible, but
this is in direct contrast to the phototropic action spectrum which is
almost at a minimum at 390 πΐμ.
While the action spectrum may thus be considered rather definitely
to exclude riboflavin as a photoreceptor, it does not necessarily exclude
all flavins. In a paper which appeared long after this review was completed, Harbury et al. (124a) present absorption spectra for a synthetic
isoalloxazine, 3-methyllumiflavin. This compound in certain nonpolar
solvents shows 2 peaks and a shoulder in the visible: in CC1 4 the peaks
fall at 447 and 474; in benzene, anisole, or mesitylene, at 450 and 476
πΐμ. The peak in the ultraviolet falls at 333-341 τημ; the ratio of its
height to that of the principal peak is about 0.8. Harbury et al. consider
that the position of the ultraviolet peak depends on hydrogen bonding
with the solvent, for in water it moves up to 369, in phenol to 370, and
in formamide to 362 τημ. Unfortunately, in such solvents the shift of the
ultraviolet peak to longer wavelengths is accompanied by disappearance of the two-peaked structure in the visible, so that it appears not
to be possible by any choice of solvent to imitate the action spectrum of
phototropism. So far only lumiflavin (6, 7, 9-trimethylisoalloxazine)
and its 3-methyl derivative have been reported to show this two-peaked
spectrum in the visible, although a few flavoproteins appear to have a
shoulder on the long wavelength side of the main peak, at around 480
m/x [see especially Theorell and Akeson's curve for crystalline "old
yellow enzyme,'' (124b)]. The very carefully purified and crystallized
glycolic acid oxidase (124c) shows no such shoulder. However, it is
always possible that further study of the flavins and their derivatives,
or of the effects of solvent and of additional solutes on the absorption
spectra, might bring to light a combination of substance and conditions
giving an absorption spectrum similar to the phototropic action spectrum. Even so, it would still remain to be shown that the substance in
question does in fact occur naturally in the plant.
The proposed role of flavins in phototropism is open to question on
other grounds. In the first place, many other substances, e.g,, eosin,
KENNETH V. THIMANN AND GEORGE M. CURRY
In the action spectrum the peak in the near ultraviolet suggests
participation by riboflavin. Quantitatively, however, it is much smaller
than one would expect if riboflavin alone were the photoreceptor, because the absorption peak in the near ultraviolet of riboflavin, and of
other common flavin derivatives, is little, if at all, lower than that in
the visible. Studies of the spectral sensitivity for the photolysis of IAA
in the presence of riboflavin or concentrated brei of etiolated pea
epicotyls give action spectra closely resembling the absorption spectrum
of riboflavin (98, 121). The curves rise sharply below 410 τημ to a
height at 390 τημ surpassing the 440 τημ maximum in the visible, but
this is in direct contrast to the phototropic action spectrum which is
almost at a minimum at 390 πΐμ.
While the action spectrum may thus be considered rather definitely
to exclude riboflavin as a photoreceptor, it does not necessarily exclude
all flavins. In a paper which appeared long after this review was completed, Harbury et al. (124a) present absorption spectra for a synthetic
isoalloxazine, 3-methyllumiflavin. This compound in certain nonpolar
solvents shows 2 peaks and a shoulder in the visible: in CC1 4 the peaks
fall at 447 and 474; in benzene, anisole, or mesitylene, at 450 and 476
πΐμ. The peak in the ultraviolet falls at 333-341 τημ; the ratio of its
height to that of the principal peak is about 0.8. Harbury et al. consider
that the position of the ultraviolet peak depends on hydrogen bonding
with the solvent, for in water it moves up to 369, in phenol to 370, and
in formamide to 362 τημ. Unfortunately, in such solvents the shift of the
ultraviolet peak to longer wavelengths is accompanied by disappearance of the two-peaked structure in the visible, so that it appears not
to be possible by any choice of solvent to imitate the action spectrum of
phototropism. So far only lumiflavin (6, 7, 9-trimethylisoalloxazine)
and its 3-methyl derivative have been reported to show this two-peaked
spectrum in the visible, although a few flavoproteins appear to have a
shoulder on the long wavelength side of the main peak, at around 480
m/x [see especially Theorell and Akeson's curve for crystalline "old
yellow enzyme,'' (124b)]. The very carefully purified and crystallized
glycolic acid oxidase (124c) shows no such shoulder. However, it is
always possible that further study of the flavins and their derivatives,
or of the effects of solvent and of additional solutes on the absorption
spectra, might bring to light a combination of substance and conditions
giving an absorption spectrum similar to the phototropic action spectrum. Even so, it would still remain to be shown that the substance in
question does in fact occur naturally in the plant.
The proposed role of flavins in phototropism is open to question on
other grounds. In the first place, many other substances, e.g,, eosin,