The raspberry Gene Is Involved in the Regulation of the Cellular Immune Response in Drosophila melanogaster

Drosophila is an extremely useful model organism for understanding how innate immune mechanisms defend against microbes and parasitoids. Large foreign objects trigger a potent cellular immune response in Drosophila larva. In the case of endoparasitoid wasp eggs, this response includes hemocyte proliferation, lamellocyte differentiation and eventual encapsulation of the egg. The encapsulation reaction involves the attachment and spreading of hemocytes around the egg, which requires cytoskeletal rearrangements, changes in adhesion properties and cell shape, as well as melanization of the capsule. Guanine nucleotide metabolism has an essential role in the regulation of pathways necessary for this encapsulation response. Here, we show that the Drosophila inosine 5'-monophosphate dehydrogenase (IMPDH), encoded by raspberry (ras), is centrally important for a proper cellular immune response against eggs from the parasitoid wasp Leptopilina boulardi. Notably, hemocyte attachment to the egg and subsequent melanization of the capsule are deficient in hypomorphic ras mutant larvae, which results in a compromised cellular immune response and increased survival of the parasitoid.


Introduction
Multicellular organisms have evolved diverse defense mechanisms against pathogenic microorganisms and parasites. Cellular encapsulation of invading parasites and endogenous tumors is a phylogenetically conserved mechanism of the innate immune response. Granuloma formation in vertebrates [1,2] and capsule formation against endoparasitoid wasps in insects [3,4] represent special forms of the cellular immune response. Furthermore, throughout the animal kingdom, these responses involve phylogenetically conserved signaling molecules and modules [5], as well as structures showing remarkable similarities, most likely as a result of convergent evolution of interactions between hosts and parasites [6]. In recent years, Drosophila melanogaster has revealed itself to be an excellent model for the investigation of granuloma formation and the encapsulation reaction.

Antibodies
The P1a and P1b antibodies recognize the NimC1 molecule on plasmatocytes, the L1a, L1b and L1c antibodies [36] react with the Atilla molecule on lamellocytes [37,38]. The antibodies were used as neat in the form of hybridoma tissue culture supernatants. The secondary antibody was goat anti-mouse CF-568 (Sigma-Aldrich), used at 1:1000 dilution.

Immune induction with parasitoid wasp
The cell-mediated immune response was induced by using the parasitoid wasp Leptopilina boulardi strain G486. 72 h old Drosophila larvae (n = 50) were exposed to 8 female wasps and incubated for 2h at 25°C. After wasp infestation, the wasps were removed, and larvae were kept at 25°C or at 29°C.

Eclosion of the D. melanogaster and L. boulardi adults after parasitization and statistics
Forty-eight hours after parasitization the Drosophila larvae were collected and washed in Drosophila Ringer's solution (7.5 g NaCl, 0.35 g KCl, 0.21 g CaCl 2 , in 1000 ml dH 2 O, pH 7.0) and viewed for encapsulated and melanized wasp eggs or a small melanized black spot under a stereomicroscope. The parasitized larvae were transferred into vials containing standard fly food. The pupae were counted and then monitored for eclosing flies or wasps.
After 2 hours immune induction, the raspberry RNAi knockdown larvae were transferred to standard fly food and kept at 25°C or 29°C for 5 days respectively.
The experiments were repeated at least three times, summing up at least 100 flies/genotype. For the statistical analysis the Student's t-test was used; the p-values of <0.05 were considered as significant.

Encapsulation assay
The larvae were dissected on 12-spot microscope slides (SM-11, Hendley Essex) 48 or 72h after parasitization. The number of the live, partially encapsulated and melanized or fully encapsulated and melanized wasp larvae in each individual Drosophila larva were counted.

Immunostaining of the encapsulated wasp eggs
Forty-eight and 72 hours after wasp infestation the larvae were dissected in multiwell glass plate, in Schneider's medium containing 5% fetal bovine serum (FBS) and 0.003% 1-phenyl-2-thiourea (PTU) (Sigma-Aldrich), washed with PBS (137 mM NaCl, 2.7 mM KCl, 6.7 mM Na 2 HPO 4 , 1.5 mM, KH 2 PO 4 , pH 7.2), fixed in PBS containing 2% paraformaldehyde and washed three times in PBS for 5 min. The samples were blocked with PBS containing 0.1% bovine serum albumin (PBS-BSA) for 20 min and incubated overnight with the corresponding primary antibody. After washing three times with PBS, the CF-568 secondary antibody was added (diluted in PBS-BSA) and incubated for 45 min. The nuclei were stained with DAPI (Sigma-Aldrich). The samples were washed three times in PBS for 5 min each, mounted in Flouromount G (SouthernBiotech) and analyzed with a Zeiss Axioscope 2 MOT fluorescence microscope or a Leica confocal LSM.

Hemocyte collection, counting and statistics
Third instar wandering larvae were bled into 30 μl Drosophila Ringer's solution containing PTU by ripping the cuticle with two fine forceps. Hemocytes were counted from at least 8 larvae of each genotype in Bürker chamber. For the statistical analysis of total hemocyte number the Student's t-test was used; the p-values of <0.05 were considered as significant.

Immunostaining of circulating hemocytes
Larvae were dissected on 12-spot microscope slides in Schneider's media containing 5% FBS and PTU, at the indicated time points. Hemocytes were incubated on microscope slides to adhere for 1 hour in humid chambers in Schneider's medium containing 5% FBS, then fixed in 2% paraformaldehyde containing PBS. The samples were washed three times in PBS for 5 min and blocked with PBS-BSA for 20 min, then incubated 1 hour with the primary antibody. After washing three times with PBS, the secondary anti-mouse CF-568 antibody (diluted in PBS-BSA) was added and incubated for 45 min. The nuclei were stained with DAPI. The samples were washed three times in PBS for 5 min each, then mounted in Flouromount G and analyzed with Zeiss Axioscope 2 MOT fluorescence microscope.

Examination of pseudopod-like cytoplasmic extensions in hemocytes
Twenty-four hours after infestation with L. boulardi, six larvae were collected from each group and bled into Schneider's media containing 5% FBS and PTU. Hemocytes were incubated on 12-spot microscope slides for 1h, fixed in PBS containing 2% paraformaldehyde and washed three times in PBS for 5 min. The samples were blocked with 0.1% BSA and 0.01% Triton X-100 containing PBS for 20 min. The actin cytoskeleton was stained with phalloidin conjugated Atto Rho6G (Sigma-Aldrich, 1:1000 final dilution in PBS-BSA) for 45 min, then washed three times in PBS for 5 min each. The nuclei were stained with DAPI. The samples were mounted in Flouromount G (SouthernBiotech) and analyzed with a Zeiss Axioscope 2 MOT fluorescence microscope and the proportion of filopodia promoting hemocytes was established. Three independent biological samples were analyzed. The Student's t-test was used to calculate the significant difference from the control samples. The p-values of <0.05 were considered significant.

Results
The ras 2 allele affects the survival rate of the Drosophila larvae after infestation We determined the rate of survival of the ras 2 mutant after L. boulardi G486 parasitization in the hypomorphic, fully viable raspberry allele in D. melanogaster-as the ras null mutants are lethal [39,40]. We found that the number of the wasps eclosed from the ras 2 mutant was significantly higher (p<0.05) from the ras 2 pupae (Fig 1A, red column) than from the Oregon-R control. The number of ras 2 pupae from which neither flies nor wasps emerged was significantly lower compared to wild type controls ( Fig 1A, green column). We observed that the number of hatching flies was relatively low both in the mutant and the control (Fig 1A, blue column). In the non-infested Oregon-R and ras 2 , there was no significant difference in the eclosion rate ( Fig 1B).
We also investigated the survival of fly versus wasp after depletion the raspberry specific mRNA transcript in blood cells. We used the msnF9mo-Gal4 [34,35] driver active in lamellocytes and lamellocyte precursors and the Hemese-Gal4 (He-Gal4) [32] driver, which is active in 80% of hemocytes including plasmatocytes and lamellocytes to silence ras by two independent ras RNAi constructs (y 1 sc Ã v 1 ; P{TRIP.HMC03250} attP2 and y 1 v 1 ; P{TRIP.JF01446} attP2) at 25°C (S1 and S2 Figs) and 29°C. (Fig 1C and 1D, S3 Fig). We monitored the eclosion of flies or wasps and found that significantly more wasps hatched from the msnF9mo-Gal4>P{TRIP. JF01446} progeny compared to the parental lines (p < 0.01) (Fig 1C). In the case of the other raspberry RNAi line (y 1 sc Ã v 1 ; P{TRIP.HMC03250} attP2}), the difference was also significant compared to the msnF9mo-Gal4/+ (p > 0.05) (Fig 1D). Similar results were obtained at 25°C using the msnF9mo-Gal4 driver (S1 Fig) or by using the He-Gal4 driver line at 25°C and 29°C (S2 and S3 Figs).

raspberry is involved in encapsulation and melanization reactions
We investigated the phenotype and number of capsules in D. melanogaster after parasitization with L. boulardi G486. We counted non-encapsulated living wasp larvae, partially encapsulated and melanized or completely melanized wasp eggs (Fig 2A) in the hemocoel of the ras 2 and Oregon-R larvae. Seventy-two hours after oviposition, we counted the total number of parasitoids in the ras 2 mutant and in the Oregon-R larvae, and we found no significant difference in the number of parasitoids, which indicates that L. boulardi G486 has no preference for depositing eggs into the ras 2 or the Oregon-R larvae. At the same time, we detected significantly higher number of living or partially encapsulated and melanized wasp larvae and significantly lower number of completely melanized eggs in the ras 2 mutant compared to the Oregon-R control after 48h (S4 Fig) or 72h following oviposition (Fig 2A). Similarly, using two different ras RNAi constructs driven by the msnF9mo-Gal4 driver, we detected a significantly higher frequency of living wasp larvae in the progenies than in the parental lines 48h (S5 Fig) or 72h (Fig 2B and 2C) after oviposition. We obtained similar results using the hemocyte specific He-Gal4 driver to knockdown ras expression (S6 and S7 Figs).

Incomplete lamellocyte adherence to the capsule in the ras 2 mutant
We studied the morphological features of the capsules formed around the parasitoid egg by staining the capsule with the lamellocyte specific Atilla antibody. In Oregon-R, the hemocytes formed a tight, continuous sheet around the parasite egg (Fig 3A) with strong melanization masking the nuclear staining, while in the ras 2 mutant the lamellocytes formed a loose network around the egg (Fig 3B), accompanied with marginal melanization 72h after infestation ( Fig  3C). We obtained similar results after RNAi silencing by driving the two different RNAi lines with the He-Gal4 driver line (Fig 3D and 3E), and with the lamellocyte specific msnF9mo-Gal4 driver ( S8 Fig). The number and morphological features of the circulating hemocytes in the ras 2 mutant Proliferation and differentiation are GTP-dependent processes in lymphocytes [41], and the IMPDH catalyzes the rate-limiting step of de novo synthesis of guanine nucleotides [24]. Therefore we tested whether the observed defect in capsule formation in the ras 2 mutant could be the result of a decreased or abnormal hemocyte count in the hemolymph. Hemocyte counts of infested and non-infested Oregon-R and ras 2 lines were determined, and the lamellocytes of the infested larvae were visualized. We found that the number of hemocytes (Fig 4A) was comparable and the number, as well as the morphological characteristics of circulating lamellocytes were the same (Fig 4B and 4C).

raspberry is involved in the formation of pseudopod-like cytoplasmic extensions in hemocytes
After immune induction, hemocytes undergo functional and morphological changes. As Rac2 has a role in these alterations [20], we tested ras 2 mutants for morphological changes using Oregon-R and Rac2 Δ as controls. In the ras 2 and ras 2 ; Rac2 Δ larvae, hemocytes attached to the microscope slide were round with smooth margins and larger in size as compared to the Oregon-R, which had many pseudopod-like cytoplasmic extensions that show actin staining 24h after the immune induction (Fig 5). We found that the hemocytes of Oregon-R larvae had significantly more filopodia (7.2 ±2.9) compared to the ras 2 (3.9±1.3) (p<0.01) and ras 2 ; Rac2 Δ (4.3±2.1) (p<0.01), but this difference was not significant compared to the Rac2 Δ (5.6±4.0). In the ras 2 ; Rac2 Δ double mutant, the proportion of the extension forming plasmatocytes was lower compared to the Rac2 Δ (Fig 5A), which suggests that ras is epistatic over Rac2. The length of the filopodia was similar in the Oregon-R, ras 2 , Rac2 Δ and ras 2 ; Rac2 Δ (1-3 μm).

Discussion
The hemocytes of Drosophila melanogaster larvae form tightly compact melanized capsules around parasitoid wasp eggs. This cellular encapsulation reaction has long been studied in D. melanogaster and in other Drosophila species [3,6], however not much is known about the regulation of this defense mechanism. Small GTPases Rac1, Rac2 and the Jun N-terminal kinase Basket were found to be required for the proper encapsulation reaction against the endoparasitoid wasp L. boulardi [20,21]. Additionally, the Drosophila βPS-integrin Myospheroid is necessary for hemocytes to adhere to the cellular capsule surrounding L. boulardi eggs, and Rac1 is required for the topographic localization of Mysopheroid in the cell membrane [42]. In a directed genetic screen, several genes were isolated that are related to encapsulation defects. The mammalian homologues of many of these genes are involved in wound healing, cellular adhesion and thrombosis [43].
In a screen [44] for genes involved in the regulation of cellular immune responses, we isolated a raspberry mutant (ras 2 ) that shows significantly decreased chances for survival of the Drosophila melanogaster host after parasitisation with the parasitoid wasp Leptopilina boulardi. The ras 2 mutant has a 5.0 kb insert in the raspberry gene [39]. We localized the site of the insertion in the second intron (data not shown). We found significantly higher number of capsules with loosely attached lamellocytes and improper melanization compared to the tight, compact and melanized capsules of the Oregon-R control. This shows that both the adhesion of lamellocytes and the melanization are affected by the mutation or, alternatively, improper lamellocyte attachment may inhibit melanization. To validate our results, we also used two independent ras RNAi constructs. We found that the two RNAi lines showed different efficiency, which may be due to differences in their genetic backgrounds. Both ras RNAi lines were able to mimic the mutant dark red eye color phenotype when driven with the lozenge-Gal4 driver (data not shown), which indicated that the RNAi lines were fully functional. We noticed similar phenotypes after knocking down the raspberry transcript in hemocytes with the msnF9mo-Gal4 and the He-Gal4 drivers similarly to Bausek and Zeidler [45] after reduction of Ga73B levels in a Hop T42 background. The lower Ga73B gene dose, compared to the wild type pretumors, prevented the formation of the tightly bound cell mass and resulted in the formation of loosely associated cell clumps, which were not able to develop and generate the melanized tumors visible in adults. These defects were not associated with a decrease in hemocyte number or abnormal lamellocyte morphological features. Ga73B is a GTP binding protein, the α subunit of the heterotrimeric G proteins and a novel JAK/STAT pathway gene involved in the JAK/STAT-mediated tumor formation. The JAK/STAT pathway is also involved in the proper encapsulation response [46,47]. The regulation of the actin-based membrane protrusions, such as lamellipodia and filopodia, require coordinated events in cytoskeletal remodeling. Central to this process are the small GTPases [8,20,48]. Incomplete encapsulation reactions, similar to those we have observed in the ras 2 mutant, were described in the Rac2 Δ mutant, in which plasmatocytes and lamellocytes adhered to the parasitoid egg fail to spread and there is a failure of melanization [20]. Rac2 is a member of the Ras small GTPase superfamily involved in the formation of filopodia and lamellipodia [49,50]. Filopodia formation of hemocytes in the ras 2 and in the Rac2 Δ mutant was defective, suggesting that similarly to the D. melanogaster Rac2 [20], myospheroid [42] and TM9SF4 [51], the Drosophila raspberry gene is also involved in the encapsulation reaction via regulation of filopodia or lamellopodia formation. The lower proportion of filopodia forming plasmatocytes in the ras 2 ; Rac2 Δ double mutant suggests that ras 2 has an epistatic effect on Rac2 Δ . Interestingly, it was shown that inhibition of the de novo GMP synthesis pathway has a strong effect on small GTPase function, including Rho GTPases such as Rac2 [52,53]. Additional evidence for the involvement of de novo synthesis of guanine nucleotides in cellular processes was found in axon guidance too [54], where the Burgundy catalyzes the final reaction of the de novo GMP synthesis, while Raspberry catalyzes the first step of the same pathway both in neurons [54] and in hemocytes.
Possible explanation for the role of raspberry in the encapsulation reaction is that it encodes the rate-limiting enzyme for GTP synthesis and thus influences the function of enzymes requiring GTP. It is known that G proteins and small GTPases are involved in the regulation of processes related to the immune response. However, further studies must be conducted to elucidate the exact pathways affected in the case of decreased IMPDH level in the encapsulation reaction.