The metabolism of histamine in the Drosophila optic lobe involves an ommatidial pathway: β-alanine recycles through the retina

Abstract

Flies recycle the photoreceptor neurotransmitter histamine by conjugating it to β-alanine to form β-alanyl-histamine (carcinine). The conjugation is regulated by Ebony, while Tan hydrolyses carcinine, releasing histamine and β-alanine. In Drosophila, β-alanine synthesis occurs either from uracil or from the decarboxylation of aspartate but detailed roles for the enzymes responsible remain unclear. Immunohistochemically detected β-alanine is present throughout the fly's entire brain, and is enhanced in the retina especially in the pseudocone, pigment and photoreceptor cells of the ommatidia. HPLC determinations reveal 10.7 ng of β-alanine in the wild-type head, roughly five times more than histamine. When wild-type flies drink uracil their head β-alanine increases more than after drinking l-aspartic acid, indicating the effectiveness of the uracil pathway. Mutants of black, which lack aspartate decarboxylase, cannot synthesize β-alanine from l-aspartate but can still synthesize it efficiently from uracil. Our findings demonstrate a novel function for pigment cells, which not only screen ommatidia from stray light but also store and transport β-alanine and carcinine. This role is consistent with a β-alanine-dependent histamine recycling pathway occurring not only in the photoreceptor terminals in the lamina neuropile, where carcinine occurs in marginal glia, but vertically via a long pathway that involves the retina. The lamina's marginal glia are also a hub involved in the storage and/or disposal of carcinine and β-alanine.

Fig. 1.

Pathways for the biosynthesis and recycling of β-alanine. (A) After its release from the photoreceptor terminal, where it acts on the dendrites of large monopolar cells (LMC), histamine (HA) is recycled by a glial shuttle pathway in the lamina that requires β-alanine (β-A) in the epithelial glia (EG) for the biosynthesis of β-alanyl-histamine (carcinine, CA) under the influence of Ebony, and that releases β-alanine along with histamine in the photoreceptor terminal (R1–R6) under the action of Tan. (B) In insects, β-alanine is synthesized from aspartate or uracil by two main pathways, regulated either from aspartate by aspartate decarboxylase (black), or from uracil (C) in succession by dihydropyrimidine dehydrogenase [Pyd1, encoded by su(r)¹], dihydropyrimidine amidohydrolase (dihydropyrimidase, Pyd2) and β-ureidopropionase (β-alanine synthase, Pyd3). Synonyms for the corresponding Drosophila genes involved in the synthesis of β-alanine from uracil are indicated. Mutant black¹ (b¹) lacks the first pathway, mutant su(r)¹r^C lacks the second pathway and double mutant su(r)¹r^C; b¹ lacks both pathways (B).

Fig. 2.

(A) Head β-alanine (β-ala) in different mutants, relative to wild-type. Mutant ebony [as the homozygous inversion In(3R)e^AFA (Caizzi et al., 1987)] has significantly more β-alanine than wild-type (wt), while tan (t¹) has significantly less, as with increasing severity do black (as b¹) and the Pyd1 Black double mutant [as su(r)¹r^C; b¹]. Differences from wild-type are significant (^*P<0.05). (B) Total head β-alanine content in wild-type, and mutant black (as b¹) and double mutant pyd1; black [as su(r)¹r^C; b¹]. Values are from control flies (CTRL) and, relative to these, from flies fed different β-alanine substrates, either 5% l-aspartate (ASP) or 5% uracil (URA). These caused head β-alanine content to increase (^*P<0.05) in wild-type and black mutant flies, but not in the double mutant.

Fig. 3.

(A) Total head histamine (HA) in wild-type (Oregon R), mutant black (as black¹) and double mutant pyd1; black [as su(r)¹r^C; b¹]. Values are from control flies and, relative to these, from flies fed either β-alanine (β-Ala) or one of two different substrates, l-aspartate (ASP) or uracil URA). In general, head histamine did not differ from the corresponding control (CTRL) values after feeding for any genotype, except for the content in mutant black¹ measured after drinking uracil and double mutant su(r)¹r^C; b¹ after drinking β-alanine, in which there was a significant increase (^*P<0.05). (B) Head β-alanine in four double mutant genotypes for black and either pyd2 or pyd3, either controls (CTRL) or flies given solutions of one of two different substrates, l-aspartate (ASP) or uracil URA), for 2 h. Uracil caused a significant increase in head β-alanine in all genotypes (^*P<0.05) while aspartate caused an increase in head β-alanine in the pyd2 single mutant (first genotype), but this was smaller than the increase with uracil. (C) Distribution of ³H in the heads of tan¹ flies after drinking 25% [³H]β-alanine for 5–45 min, measured as radioactive scintillations (counts min^–1, CPM) in individual 1 min HPLC fractions. A clear peak corresponding to the retention time of β-alanine (β-ala) increases with the time during which flies could drink from the radiolabelled solution. A small peak of ³H-labelled carcinine (CA) appears at a retention time of 13 min in the chromatogram. After 30 min of drinking ³H-labelled β-alanine the carcinine peak was 22.6 CPM, 2.03 times more than the mean baseline rate for 10–16 min (less the 13 min peak) of 11.12 CPM. After 45 min drinking, the peak was 27.4 CPM, 1.34 times higher than the corresponding baseline level. These peaks are to be compared with those in flies that drank [³H]histamine, as previously reported (Borycz et al., 2002). (D) Total head content of histamine (HA) in wild-type and tan¹ flies fed diaminobutyric acid (DABA) at two different concentrations relative to controls (CTRL). No changes were seen in wild-type flies (Oregon R), but in tan both concentrations of DABA produced significant increases in head histamine (^*P<0.05).

Fig. 4.

β-Alanine concentration affects the size of the ‘off’ transient of the electroretinogram (ERG). ERGs were recorded for wild-type OR, a double mutant of the uracil-based pathway of β-alanine production su(r)¹r^C and the triple mutant su(r)¹r^C; b¹, which affects both β-alanine production pathways. (A) Representative ERGs averaged from 10 light flashes at 20 s intervals for OR, su(r)¹r^C and su(r)¹r^C; b¹ females after they drank a solution of 4% glucose or 5% β-alanine in 4% glucose. The ‘on’ and ‘off’ transients, as well as the sustained negative response (SNR), are indicated on the trace for glucose-fed OR flies. (B) Mean normalized ‘on’ and ‘off’ transients for OR, su(r)¹r^C and su(r)¹r^C; b¹ flies. Error bars: ±1 s.d. Differences are significant (^*P<0.05) within (black) and between (grey) treatments for both ‘on’ and ‘off’ transients.

Fig. 5.

Distribution of immunoreactivity to β-alanine in the heads of wild-type and mutant Drosophila. (A) Wild-type Oregon R shows label in the photoreceptors, especially in the lamina (arrowhead), a layer beneath the cornea (arrow) and also along the basement membrane region (double arrowhead). (B) Wild-type Oregon R after drinking 0.5% histamine shows increased label in the photoreceptor cell bodies (arrow), especially in the sub-corneal (arrowhead) and basement membrane (double arrowhead) layers. (C) Mutant In(3R)e^AFA shows a weaker signal in the subcorneal layer and a somewhat stronger signal in the retina compared with wild-type (in A). (D) Mutant tan¹ shows a pattern like that of In(3R)e^AFA (in C), but with a weaker band of label at the level of the basement membrane. (E) Mutant black¹ and (F) double mutant pyd1; black¹ show no clear differences from each other or from ebony¹ [In(3R)e^AFA] and tan¹. R, retina; L, lamina; M, medulla. Scale bar, 100 μm for all panels.

Fig. 6.

(A) Immunoreactivity to β-alanine (magenta) and its distribution in identified cells and organelles of the wild-type Drosophila retina and lamina. (A) Photoreceptor cells in the retina labelled by expression of β-galactoside driven by rh1-lacZ (green). Compared with β-alanine, there is faint expression in photoreceptors (arrows in A″) but colocalization is mostly lacking, with β-alanine filling the gaps (arrows) between rh1-driven β-gal expression in the photoreceptors. (B) Pigment and cone cells in the retina, labelled by expression of β-galactoside driven by spa-lacZ (green), which overlaps extensively with β-alanine immunosignal in pigment cells (B′). (B″) Regions of overlap are extensive (arrows). (C) Immunoreactivity to VMAT-B (green) labels the fenestrated glia beneath the basement membrane, which overlaps with immunoreactivity to β-alanine (C′). (C″) Signal overlaps (arrows). Overlap is less clear in a region of β-alanine expression above the basement membrane that corresponds to the location of the pigment cell end feet. (D) Photoreceptor terminals R1–R6 in profiles of cross-sectioned lamina cartridges revealed by expression of β-galactoside driven by rh1-lacZ (green). Immunolabelling for β-alanine reveals signal of about equal strength in both the terminals of R1–R6 and the surrounding epithelial glial cells (D′). (D″) The brightest β-alanine puncta (arrow) lie mostly at the borders of R1–R6. Scale bar in D″, 10 μm for all A, B and D panels; scale bar in C″, 20 μm for all C panels.

Fig. 7.

Distribution of immunoreactivity to carcinine in the heads of wild-type and mutant Drosophila. (A) Wild-type Oregon R shows label in a layer beneath the cornea (arrow) and also in the marginal glia (filled arrowhead). The subcorneal layer probably includes primary pigment cells because slender immunolabelled profiles (open arrowhead) extend down to the basement membrane. Signal also appears in a subretinal layer. (B) Mutant ebony¹ shows a reduction in signal in all parts of the brain, noticeable here in the optic lobe relative to the wild-type (in A). (C) Mutant tan¹ shows label in the layers beneath the cornea (arrow) and basement membrane (arrowhead) that is stronger than in the wild-type; immunosignal in the subretinal layer probably arises from the fenestrated glia. Marginal glia are not labelled. Scale bar, 100 μm for all panels.

Fig. 8.

Carcinine (green) and β-alanine (magenta) in the wild-type retina colocalize in primary pigment cells and photoreceptors. (A–A″) In longitudinal sections of the retina, the brightest signal for both epitopes is observed in the primary pigment cells (A″, arrow), shown enlarged in B–B″ (arrowheads in B″). In addition, β-alanine signal also occurs in the pseudocone cavities (arrow in B′). (C–C″) In cross-sections of ommatidia, additional sites of β-alanine signal are visible in the secondary and tertiary pigment cells that surround the ommatidium (arrows in C′), and colocalization with carcinine immunoreactivity is apparent in the photoreceptors (C″). Scale bar, 20 μm in A′ for all A panels; 10 μm in B′ for all B and C panels.

Fig. 9.

Cellular localization of β-alanine (β-A, blue), carcinine (CA, magenta) and histamine (HA, green) in cells of the ommatidium and underlying lamina cartridge. At least some carcinine is proposed to diffuse in a retrograde direction back into the photoreceptor cell body (magenta arrow). β-Alanine is mostly found in the pseudocone and the pigment cells that surround it, and in fenestrated glia, where it may derive from the end feet of the pigment cells. Carcinine is clearly detectable in marginal glia, where it most likely derives from the neighbouring epithelial glia, which are Ebony expressing and therefore produce carcinine. β-Alanine occurs in a subretinal layer that probably contains the fenestrated glia and could provide a pathway for β-alanine to return to the lamina glia. Such a tentative pathway would depend on the presence and patency of gap junctions or of appropriate transporters both between the pigment and glial cells and among the glial cells. Alternatively, β-alanine may support some quite different function locally in the retina. Abbreviations: L-H, l-histidine; HDC, histidine decarboxylase; l-aspartate, D.