Preventing Neurodegeneration in the Drosophila Mutant bubblegum

See allHide authors and affiliations

Science  18 Jun 1999:
Vol. 284, Issue 5422, pp. 1985-1988
DOI: 10.1126/science.284.5422.1985


The Drosophila melanogaster recessive mutantbubblegum (bgm) exhibits adult neurodegeneration, with marked dilation of photoreceptor axons. The bubblegummutant shows elevated levels of very long chain fatty acids (VLCFAs), as seen in the human disease adrenoleukodystrophy (ALD). In ALD, the excess can be lowered by dietary treatment with “Lorenzo's oil,” a mixture of unsaturated fatty acids. Feeding the fly mutant one of the components, glyceryl trioleate oil, blocked the accumulation of excess VLCFAs as well as development of the pathology. Mutant flies thus provide a potential model system for studying mechanisms of neurodegenerative disease and screening drugs for treatment.

Drosophila, with its highly evolved nervous system, short generation time, and amenability to genetic and molecular analysis, offers an incisive tool for the analysis of hereditary neurodegenerative diseases (1). For example, the fly mutants swiss cheese (2),spongecake, and eggroll (3) show late-onset neurodegeneration in the adult brain resembling that seen in various human diseases.

In screening for Drosophila brain degeneration by P-element insertion mutagenesis (4), we isolatedbubblegum, a mutant gene on chromosome 2. Upon cloning the gene, we found that it was similar to vertebrate VLCFA acyl coenzyme A (CoA) synthetase, suggesting possible relevance of the fly mutant to human ALD, which is associated with excess VLCFAs due to mutations in a transporter gene for that enzyme. We therefore examined the VLCFA profile in the fly mutant and found that it was also abnormal.

The development of neuropathology with age in the visual system of bubblegum flies was examined by light and electron microscopy (Fig. 1) (5). In the young fly mutant, the optic lobes appear normal but, with age, regionally specific degeneration develops that is not seen in the parental strain. This is particularly marked in the first optic ganglion, the lamina (corresponding to the vertebrate eye ganglion cell layer), in which photoreceptor axons enter to synapse with second-order neurons. The appellation bubblegum refers to the bubbly appearance of the lamina, as seen in light microscope sections (Fig. 1B). There is also degeneration of the cell bodies in the retina. Electron micrographs show inflation of various structures, which is most evident in the expansion in the diameter of the photoreceptor axons (Fig. 1, C and D). This pathology is similar in male and female homozygous bubblegum flies.

Figure 1

Optic lobe degeneration in bubblegummutant flies. (A and B) Shown are horizontal sections of plastic-embedded adult male heads, stained with toluidine blue. (A) Young mutant fly (1 day old). (B) A 15-day-old mutant. Note bubbly appearance of the lamina. La, lamina; Re, retina. Bar, 50 μm. (C and D) Electron micrographs. (C) Lamina of 1-day-old bubblegum male, sectioned parallel to the photoreceptor axons (arrows). The axon diameter (indicated by brackets) is the same as in normal flies (not shown). (D) At 15 days of age, the mutant photoreceptor axons have become greatly dilated. Bar, 2 μm.

To identify the affected gene, we isolated DNA adjacent to the P-element insertion by plasmid rescue (6). The genomic sequence corresponded to that located at position 34F1-2 on chromosome 2, as determined by the Berkeley Drosophilagenome project (, and a cDNA clone was obtained from Genome Systems (St. Louis, Missouri). Analysis of the mutant located the P-element insertion between the 5′ untranslated region (5′UTR) and the start codon of an open reading frame of 1947 base pairs (bp) that predicts a protein of 649 amino acids. A BLAST search indicated homology with a VLCFA acyl CoA synthetase expressed in rat liver (7) and human brain (8) (Fig. 2A). Thebubblegum gene is the only known VLCFA acyl CoA synthetase–like gene in Drosophila.

Figure 2

The bubblegum gene and metabolic pathway for VLCFAs. (A) Alignment of thebubblegum protein with rat VLCFA acyl CoA synthetase (Rat VLCFA Synth.) and a human brain protein sequence of as yet unidentified function. Identical and similar amino acids are identified by dark and light shading, respectively. Compared with the rat synthetase,bubblegum has 15% identity and 32% overall similarity, but the homology with the human sequence is greater: 40% and 61%, respectively. (B) Genomic map of thebubblegum gene region. Triangle indicates the P-element insertion. Open and hatched rectangles indicate untranslated and coding sequences, respectively. The exon-intron structure of the 3′ region has not been determined. Restriction enzyme site: H, Hind III; K,Kpn I; N, Nco I; S, Sal I; X, Xho I. (C) Transcription of thebubblegum gene is greatly reduced in the mutant fly. Reverse transcription–polymerase chain reaction, using 1 μg of total RNA from the parental strain (lanes 2 through 4) or bubblegum(lanes 5 through7), was used to detect the bubblegumtranscript. The two primers used were chosen from the 5′UTR and 3′UTR sequences. Lanes 2 and 5, whole fly; lanes 3 and 6, bodies only; lanes 4 and 7, heads only. (D) Normally, the synthetase activates VLCFAs for degradation. In ALD patients, a genetic defect in an ABC transporter interferes with the function of the synthetase. As a consequence, there is an excess accumulation of VLCFAs. In thebubblegum mutant, a similar effect is due to a mutation in the synthetase itself.

This enzyme activity is reduced in X-linked human ALD (9), which is manifested at a young age by neural demyelination and blindness. In a milder form, referred to as adrenomyeloneuropathy (AMN), onset is later, with progressive paraparesis and distal axonopathy (9). The β-oxidation of VLCFAs is catalyzed in peroxisomes, in several steps, after activation to thioester derivatives by the synthetase (10, 11). Reduced activity of the enzyme can result in accumulation of excess VLCFAs (12), as can a defect in any step of the oxidation process (13). ALD and AMN patients have mutations not in the gene encoding VLCFA acyl CoA synthetase, but in that for a member of the ABC transporter superfamily (ALDP), which is thought to be needed for the transport or stabilization of the enzyme (14). According to the pathway (Fig. 2D), mutations in the synthetase gene, as in bubblegum, would be expected to cause the same disturbance as those affecting the transporter.

To study bubblegum, we analyzed fatty acids from 15-day-old whole flies by gas chromatography (15), comparing homozygous mutant flies with the parent strain. Mutant males indeed showed increased VLCFAs, including C22, C24, and C26, whereas those of chain length below C20 did not increase (Fig. 3A). Paradoxically, although both male and female mutant flies showed similar neuropathology, homozygous females showed little change in VLCFA levels, presumably because of differences in the metabolism of VLCFAs (Fig. 3B).

Figure 3

Preventing the accumulation of VLCFAs in bubblegum males. Profile of adult bubblegum(A) males and (B) females raised for 15 days in ordinary medium. (C) Adult males that were raised in medium containing 2.5% GTO from the first larval instar through to 15 days of adult age. Each bar represents the ratio of the homozygous bubblegum mutant to the parental strain, based on four experiments, each using 100 flies. Standard deviations are shown.

In human ALD, dietary treatment has been used to restore VLCFAs to normal levels. Feeding patients a monounsaturated fatty acid, glyceryl trioleate oil (GTO), reduced the level of saturated C26 in plasma by about 50% within 4 months (16). So-called “Lorenzo's oil,” a combination of GTO with glyceryl trierucate oil, normalized C26 in a month (17). These monounsaturated fatty acids probably compete in the fatty acid elongation system, reducing the formation of the saturated VLCFAs (18). We therefore tested whether similar dietary treatment can reduce the concentration of C26 and prevent the pathology seen in bubblegum.

When larvae were raised on the usual cornmeal-yeast-agar medium and the emerging adults transferred to medium supplemented with 2.5% GTO, the lamina was still somewhat abnormal (Fig. 4A). Considering that oil treatment in ALD could not prevent progression of the disease once patients had shown neurological symptoms (17), we tried beginning treatment with the oil at preadult stages, continuing into adulthood. The amounts of VLCFAs in male flies were reduced to normal (Fig. 3C). In addition, degeneration was prevented (Fig. 4B), both in male and female mutant flies. At the temperature used (29°C to accelerate aging), the average life-span was 19 days for the parental strain and 13 days for bubblegum; GTO treatment restored the life-span of bubblegum to that of normal flies. The oil had no evident deleterious effect on normal flies.

Figure 4

Preventing degeneration in bubblegumby early initiation of feeding with GTO. Sections and staining are as in Fig. 1. (A) Late initiation. Homozygous malebubblegum that were placed in GTO-containing medium for 15 days following their emergence as adults still showed considerable degeneration in the lamina. (B) For homozygous malebubblegum that were raised in GTO-containing medium from the larval stage onward through to 15 days of adult age, the late-onset adult degeneration was largely prevented. Bar, 50 μm. (C) Rescue of phototactic behavior of bubblegum by feeding GTO. In the multiple-trial countercurrent distribution experiment, the ordinate represents flies that responded positively the number of times indicated by the abscissa, out of five trials. Flies raised without GTO (open circles) showed poor performance. Flies pretreated with GTO from the larval stage (solid circles) showed greatly enhanced responses.n = 100 bubblegum adult males, 5 days of age, for each graph.

In ALD, visual loss occurs (9), and anatomical analysis has shown abnormalities of the optic nerve and degeneration of the ganglion cell layer (19). We therefore tested the vision of bubblegum. Flies transformed with P-elements marked with the mini-white + gene, as inbubblegum, express some eye pigment and are phototactic (20). We used the countercurrent phototaxis test, in which a population of flies is fractionated according to number of positive responses in repeated trials in movement toward light (21). As seen in Fig. 4C, the mutants raised in ordinary medium showed poor response (flies responding in 10% of all trials), consistent with their visual system abnormalities, whereas those raised with GTO from the larval stage onward showed greatly improved phototaxis (responding in 60% of all trials).

The difference between male and female mutant flies homozygous for the second chromosome P-element insertion is curious. Females showed the same age-dependent degeneration, which was prevented by feeding the oil, despite not showing the excess VLCFAs seen in males. The mutation in the synthetase is sufficient to produce the pathology, but not to cause a high level of VLCFAs in both sexes, implying that other, sex-related factors influence the accumulation of VLCFAs.

In X-linked human ALD (9) and in three knockout mice of the ABC transporter gene (22–24), there was no correlation between the amount of VLCFAs and the severity of pathology. Some individuals with high VLCFAs escaped the neurological defects (9), and the knockout mice did not show the pathology of ALD, in spite of having elevated VLCFAs. These observations suggest that excess VLCFAs and the pathology may not have a direct causal relationship, but may be separate ramifications of another, underlying defect. H. W. Moser et al.(9) suggested that autosomal modifer genes may play a role in the pathogenesis of ALD. The differences in our observations between males and females are consistent with that interpretation. InDrosophila, it should be possible to do genetic analysis to clarify this problem by identification of suppressor and enhancer genes.

The majority of human ALD mutations have been traced to the transporter gene. The fact that mutations in the synthetase gene inbubblegum produce effects that have some features in common with ALD suggests that additional screening of patients and their families for mutations in the human synthetase gene might be desirable. Knockout of the transporter gene in mice has so far not resulted in an ALD phenotype (22–24). It remains to be seen whether knockout of the synthetase or double knockout of both the enzyme and the transporter will do so.

The pathological phenotype in bubblegum is prevented by feeding the oil. Although ALD patients treated with “Lorenzo's oil” have experienced a reduction in the amount of C26, there has been little success in preventing progression of the disease, perhaps due to the late start of treatment, which also failed to rescuebubblegum. Because experiments on optimization of treatment are difficult to perform on human subjects, it is possible that fly mutants such as bubblegum can serve as useful model systems for rapidly screening food additives, drugs, and regimens of administration.


View Abstract

Stay Connected to Science

Navigate This Article