Lipid droplet distribution is altered in pummelig mutant flies

As introduced before, body fat is increased in puml¹ and lipid storage is elevated in Malpighian tubules that can be rescued by tissue-specific expression of puml-cDNA.

Microscopic analysis of lipid storage using fluorescent lipophilic dyes indicated a difference in the lipid droplet size distribution between puml¹, bmm¹ and control

flies. Due to the limited biological dimensions of Malpighian tubules, this tissue was used to analyse the lipid droplet size between these genotypes (Figure 32).

Figure 32 Average lipid droplet size (diameter) is reduced in Malpighian tubules of puml¹ flies. Box plot of lipid droplet size quantified from confocal pictures of fluorescently stained lipid droplets in Malpighian tubules. Center lines show the median, box limits indicate 25^th and 75^th percentiles as determined by OriginPro software; whiskers extend 1.5 times the interquartile range from the 25^th and 75^th percentiles (Mann-Whitney test; nlipid droplets analysed per genotype: bmm¹=1206, puml¹=1715 and control=306).

Lipid storage is quite low in Malpighian tubules of control flies (Figure 32) reflected by the low number of individual lipid droplets that were used for the size distribution (diameter of lipid droplets was calculated from the total area from a maximum intensity projection of an acquired 3D-confocal image stack). Though, the quantified lipid droplets on average had the largest diameter compared to bmm¹ and puml¹. As described earlier both mutants had a much higher lipid storage in Malpighian tubules but average lipid droplet (LD) sizes were significantly smaller. A possible explanation might be an alternated phospholipid metabolism in puml¹ and bmm¹ flies that finally effects also the phospholipid composition of the LDs and therefore shaping their size.

3.9.2 Long-chain fatty acids and poly-unsaturated fatty acids are elevated in pummelig mutant flies

Lipid stores are elevated in puml¹. The major neutral lipid class contributing to this are TAGs (Figure 14). During starvation lipids are mobilized in general but on average a significantly higher amount of lipids remain in dead animals (Figure 21).

Additionally, lipid droplet size distribution in Malpighian tubules is changed in puml¹.

AtABHD4/5 null mutants (plants) exhibit similar to puml¹ ectopic lipid storage in leaves (James et al., 2010). The major elevated neutral lipid class were TAGs and content of poly-unsaturated fatty acids (PUFAs) esterified in the neutral lipids were elevated (James et al., 2010) and a shift towards longer fatty acid sidechains of neutral lipids could be seen. Additionally, it had been shown that AtABHD4/5 interacts with the peroxisomal ABC-transporter 1 (PXA1) (Park et al., 2013). A single PXA1 mutant and a AtABHD4/5, PXA1 double mutant exhibited the same phenotype on TAGs as the single AtABHD4/5 mutant.

As shown before Puml is localized on lipid droplets and Puml::mCherry can be found on peroxisomes as well. Also, a portion of neutral lipids remains immobilized during starvation to death in puml¹. Therefore, I hypothesized a comparable phenotype of changed TAG composition. In order to identify changes, the lipidome was analysed using mass-spectrometry. Malpighian tubules were explanted from adult puml¹ and control flies on standard diet (density controlled). Lipid extraction and mass spectrometry were performed by Vinzenz Hofferek (pilot experiment; 5 independent extractions from each 10 MT pairs; MPIMP-golm; average standard deviation of Mol

% TAG species=0,15% for both genotypes) and independently by Dr. Thomas Eichmann (University of Graz; 3 independent extractions from each 100 MT pairs;

average standard deviation of Mol % TAG species=0,50%).

The main TAG species in control flies are: 46:1, 48:0, 48:1, 50:1, 52:0 and 52:1. In puml¹ flies: 48:2, 50:1, 50:2, 52:2, 52:3 (Figure 33, Figure 34). As no total hydrolysis was performed, ratios for the single free fatty acids were not acquired and the individual composition of each TAG species can only be predicted with an amount of uncertainty. Based on the desaturation grade found in the most common lipids and their size (first number represents total number of carbon atoms found in the sidechains of the glycerolipids; second number indicated the number of C=C-bonds [desaturation grade]) it is very likely that oleic acid (C18:1) and linoleic acid (C18:2) represent the major fatty acids contributing to these TAG species.

Compared to puml¹ the most abundant TAG species had a shorter length in control flies (Figure 33, Figure 34). In correlation with a higher relative abundance of long

chain fatty acids (LCFA) also the relative abundance of TAGs with a higher desaturation grade was increased in puml¹. Both experimental approaches confirmed a higher abundance of TAGs (in absolute amounts) in Malpighian tubules of puml¹ and complement the data from microscopic analyses and TLC results of lipid extracts deriving from total flies.

The current lipidomics data also strengthens the idea of Puml being involved in the peroxisomal directed β-oxidation of long chain fatty acids (LCFAs: C13 - C21).

Comparable to AtABHD4/5 mutants we have increased abundance of PUFAs and LCFAs in puml¹. A similar phenotype can also be observed in human patients of the Zellweger syndrome (Suzuki et al., 1996).

Figure 33 LCFA-TAGs and abundance of PUFAs are elevated in Malpighian tubules of puml¹ flies.

Plotted are the means of Mol % ± SEM (n=4 independent extractions) per molecular TAG species (upper numbers represent number of C=C bonds in the fatty acid side chain; lower number the total number of carbon atoms from the esterified FAs; data from control flies are shown in blue; puml¹ data depicted in red). (A, B) Lipid extraction, mass spectrometry and lipid annotation performed by Vinzenz Hofferek (MPIMP-golm). (C, D) Lipid extraction, mass spectrometry and lipid annotation performed by Dr. Thomas Eichmann (University Graz). Long chain fatty acids (LCFAs) [50, 52] were elevated in puml¹ compared to control flies in both experiments. Additionally, more poly-unsaturated fatty acids (PUFAs) could be detected in puml¹ [48:2, 48:3, 50:3, 50:4, 50:5, 50:2, 50:3, 50:4, 50:5, 52:2, 52:3, 52:4, 52:5).

The Zellweger syndrome is characterized by a general loss of peroxisomal functions (Jones et al., 1992). This includes the impaired degradation of mono- and poly-unsaturated fatty acids (Christensen et al., 1986, Wanders et al., 1987).

Taken together, in vitro assays revealed no esterase activity for Puml on neutral lipids.

In puml¹ flies body fat is increased and TAGs mostly contribute to this phenotype (Figure 14). Multiple lipidomics analyses confirmed increased TAG storage and revealed PUFAs and LCFAs being more abundant in puml¹. This is very like caused by an involvement of Puml in the peroxisome targeted lipid degradation that at least seems to be decreased in puml¹ and would explain the dual localization on LDs and peroxisomes.

Additionally, Puml is a potent phospholipase and shows some affinity for PA, PG or BMP(R,R) suggesting a modulating role on the phospholipid monolayer of lipid droplets. This may also provide some explanation why the lipid droplet size distribution is changed in puml¹ as well.

Figure 34 Heat map of TAG species distribution shows increased abundance of PUFAs and shift towards longer fatty acid sidechains in Malpighian tubules from puml¹ flies. Lipid extraction, mass spectrometry and Lipid annotation done by Vinzenz Hofferek (MPIMP-golm; Pilot experiment;

represented are data from 5 discrete extractions from 10 Malpighian tubules [MTs] pairs) and Dr.

Thomas Eichmann (University Graz; represented are data from 3 independent extractions from each 100 MT pairs; data with n.a-coloring were not annotated).

Looking at the organismal level absence of puml¹ is not crucial as homozygous null mutants are viable. On the other hand, carbohydrate stores (glycogen) and lipid stores are modulated differently in puml¹. Whereas food intake is not impaired in

puml¹, ingested food is preferentially stored in form of TAGs to the account of carbohydrates. This may contribute to the findings that puml¹ have a higher starvation resistance but are more susceptible to desiccation.

Lipid mobilization in general is not impaired in puml¹ and a direct enzymatic activity on neutral lipids could not be detected. Therefore, the biological role of puml seems to be more indirect in assisting lipolysis, directing fatty acids and acting more like a metabolic switch at the junction between carbohydrate and lipid metabolism in Drosophila melanogaster. This would also characterize the core lipid mobilization module of ABHD5-ATGL more as a specialization during evolution providing an additional level of control and enhanced efficiency of lipid mobilization in higher organisms rather than an evolutionary conserved theme.

4 Discussion