Introduction

Tetraspanins are abundantly expressed proteins characterized by four transmembrane domains delimiting two extracytoplasmic loops of unique sizes that define a large, evolutionary conserved family (Hemler, 2001). First discovered as human leukocyte antigens, they were soon found in a variety of other tissues (Boucheix and Rubinstein, 2001). Despite numerous biochemical studies and a few characterizations of mutant phenotypes, the specific function of tetraspanins remains largely undefined. They are known to physically interact with one another or with various membrane receptors and intracellular signal mediators to form the so-called microdomain complexes. The presumption is that these complexes function as dynamic organizers in the regulation of specific signal transduction events (Yunta and Lazo, 2003) and modulate cell motility, cell aggregation and cell signaling.

To date, 20 out of 28 human tetraspanins have been shown to be expressed in blood cells (Scherberich et al., 2002; Tarrant et al., 2003). Their precise function has been difficult to pin down, as there is significant functional redundancy among the various members. Knockout studies for different murine tetraspanins showed no effect on development, but often led to a mild alteration of lymphocyte proliferation and motility (Miyazaki et al., 1997; Scherberich et al., 2002; Tarrant et al., 2002). There are 37 predicted tetraspanin genes in Drosophila (Todres et al., 2000). Of these, only two genes – late-bloomer (Kopczynski et al., 1996; Fradkin et al., 2002) and sunglasses (Xu et al., 2004) – have been characterized in detail at the functional level. Mutations in these two genes did not impair viability and caused only mild phenotypes. Furthermore, a homozygous deletion of a genetic cluster that contained about one-quarter of all tetraspanin genes in flies was surprisingly well tolerated, and resulted only in a transient defect in neuromuscular innervation during the larval stage (Fradkin et al., 2002).

Although the role of tetraspanins in blood cell function has been actively evaluated in mammals, there are no reports about their functional role in hemocytes, the Drosophila blood cells. Here, we provide evidence that the tetraspanin tsp68C gene is normally expressed in hemocytes. Furthermore, its overexpression in these cells can abrogate severe blood defects observed in ytr mutants or in larvae expressing activated forms of Raf or Ras in hemocytes. Our results also illustrate the value of taking a gain of function approach as it overcomes potential problems of redundancy inherent to loss of function studies with tetraspanins. This alternative approach may prove highly informative in deciphering the function of these proteins not only in Drosophila but also in other organisms.

Results

An hml-Ga4 construct restores normal hemocyte differentiation and proliferation in ytr mutants

Recently, we reported that mutations in the ytr gene, encoding an arginine-rich nuclear protein resulted in animal death during late larval to early pupal phase. The mutations were associated with growth defects in imaginal discs and with abnormal hemocyte proliferation and differentiation in third instar larvae (Sinenko et al., 2004). To examine whether the effect on hemocytes was cell autonomous, we embarked on experiments to re-express the wild-type ytr gene in the hematopoietic tissue of mutant larvae. We used the dual expression system based on the UAS/Gal4 technique (Brand and Perrimon, 1993) and selected hml-Gal4 transgenic flies to drive expression of UAS-ytr. These flies carry a synthetic construct consisting of the 3.0 kb region upstream of the hemolectin gene fused to the coding sequence of Gal4 (Goto et al., 2001). There is no significant Gal4 expression in these lines until the start of the larval stages when Gal4 is expressed exclusively in the larval lymph gland (the larval hematopoietic organ) and circulating hemocytes (Goto et al., 2001). To facilitate the rescue experiment, we recombined the hml-Gal4 construct with two different mutant alleles of ytr, ytr^5E24 and ytr^P22. As expected, the recombined stocks failed to complement the original ytr mutants and homozygous larvae from these stocks exhibited the characteristic imaginal disc defects and lethality associated with ytr mutations. Surprisingly, homozygous larvae from all recombined stocks showed a near-complete rescue of the hemocyte and lymph gland defects, even though we had not provided a copy of the wild-type ytr gene in these cells (Figure 1). The majority of circulating hemocytes showed normal plasmatocyte morphology, and hemocyte counts were comparable to those of wild type or heterozygous larvae. To rule out a possible artifact resulting from the recombination event used to derive these stocks, we generated a ytr^5E24 mutant stock in which the hml-Gal4 construct was present on a different chromosome (chromosome III) from where the ytr gene resides (chromosome II). Larvae that were homozygous for the ytr^5E24 mutation and carried one copy of the hml-Gal4 construct on the third chromosome behaved identically to the previous stocks and showed near complete abrogation of the hematopoietic defects (Figure 1). Thus, even a single copy of the hml-Gal4 construct is sufficient to restore normal blood in ytr mutants. Gal4 expression alone in hemocytes does not restore normal function to ytr mutant hemocytes. Indeed, homozygous ytr mutant larvae, carrying two copies of the daughterless-Gal4 construct (which drives ubiquitous expression of Gal4, including in the larval hematopoietic tissue) showed no significant rescue of the characteristic defects in lymph gland and circulating hemocytes (Sinenko et al., 2004). Searching for an explanation, we noticed that the 3.0 kb promoter region in the hml-Gal4 construct contained a single long open reading frame (ORF), which had been originally overlooked in the first Drosophila genome annotation despite being predicted by Todres et al. (2000). This ORF codes for Tetraspanin 68C (Tsp68C), an evolutionary conserved protein with homology to members of the tetraspanin superfamily of proteins (Figures 2a and 5a). RT–PCR performed on mRNA isolated from whole larvae and larval hemocytes confirmed that tsp68C was expressed in wild-type animals (wt), in particular in hemocytes, and that its expression was dramatically elevated in hml-Gal4 and ytr^5E24, hml-Gal4 third instar larvae (Figure 2b and data not shown). The increase was primarily due to the contribution of transcripts driven by the hml promoter in hemocytes (compare hml-Gal4 larvae vs hml-Gal4 hemocytes, Figure 2b). Interestingly, the level of tsp68C expression is higher than expected from an additive effect of having two extra tsp68C genes in the hml-Gal4 lines, suggesting that negative element(s) in the tsp68C regulatory sequence are missing in the hml-Gal4 construct. Taken together, these data suggest that the rescue of the hematopoietic defects in ytr mutants strictly correlated with the presence of the hml-Gal4 construct and likely resulted from the high level of tsp68C expression associated with this construct.

Figure 1.

Rescue of the differentiation and proliferation defect in ytr mutant hemocyte by the hml-Gal4 constructs. (a) Hemocytes from a single larva of each genotype were collected in 50 l of Ringer solution in a single well of a 10-well glass slide (PGC Scientifics). Cells were let to adhere before fixation with a 2% formaldehyde solution (Sigma). Multiple (10) independent samples for each genotype were analysed to ensure reproducibility. (b) To determine total hemocyte numbers in different mutants, the hemocytes from five larvae of each genotype were collected in 50 l of Ringer solution. The resuspended cells were counted using a hemocytometer and counts were used to determine the number of cells per single larva

Full figure and legend (173K)

Figure 2.

The hml promoter in the hml-Gal4 construct contains a tetraspanin gene, tsp68C, and drives its overexpression in hemocytes. (a) Schematic representation of the 3.0 kb hemolectin promoter region cloned into the hml-Gal4 vector. There are two forms of the tsp68C mRNA transcripts: a spliced (RA) and long unspliced (RB) form. Positions of unique KpnI site and early stop codon (see below) are indicated with arrows. (b) tsp68C transcripts from all tissues or hemocytes of control (w larvae) and hml-Gal4 larvae were estimated after amplification with the one-step RT–PCR reaction kit (Invitrogen). Ribosomal protein Rp49 transcripts were used as loading controls for all samples. (c) Position in the tsp68C second exon sequence where a TAG stop codon was introduced to interrupt the Tsp68C ORF in hml^Stop-Gal4. A HindIII site was added immediately after the stop codon to facilitate identification of the transcripts from the new transgene after RT–PCR. The KpnI/NotI fragments corresponding to the early stop codon (hml^Stop) or truncated hml promoter (hml) were cloned upstream of Gal4 ORF in the CaSpeR4 vector. (d) Hemocyte-specific expression of Gal4 in hml^stop-Gal4 and hml-Gal4 larvae was determined by crossing these stocks with UAS-GFP transgenic flies. Hemocyte preparations and its images (63 ) were taken as described in Materials and methods

Full figure and legend (216K)

Figure 5.

tsp68C overexpression specifically suppresses abnormal Ras/Raf signaling in hemocytes. (a) Hemolymph from third instar larvae resulting from crosses between a UAS-cRaf1^gof stock (Brand and Perrimon, 1994) and the modified hml-Gal4 lines, (hml^stop) and (hml) or the original hml-Gal4 line, hml(II) (Goto et al., 2001) was visualized as described above. (b) Hemocyte counts are plotted for third instar, wild-type larvae (Ore-R) and for larvae resulting from crosses between UAS-cRaf1^gof and the original hml-Gal4 lines, hml(II) and hml(III), or the modified hml-Gal4 lines, hml^stop and hml The genotype or the identity of the hml-Gal4 line is indicated under each bar. Cell counts were determined as described in Figure 1b. (c) Hemocyte counts in third instar larvae resulting from crosses between UAS-Ras^85D.V12 transgenic flies (Asha et al., 2003) and various hml-Gal4 lines, as indicated, are plotted for each condition. (d) Immunostaining of human cRaf1^gof protein in hemocytes isolated from crosses detailed under (b). Abnormal differentiation (e) and proliferation (f) caused by ectopic expression of Hop^Tum-l in hemocytes are not affected by tsp68C overexpression associated with the original hml-Gal4 line (hml(II)). Analysis of the hemolymph and cell counts were performed as described for panels a and b. Examples of lamellocytes are indicated with an arrow

Full figure and legend (366K)

Tsp68C overexpression restores near-normal hemocyte differentiation and proliferation in ytr mutants

Although we had failed to detect any other significant ORF in the 3.0 kb region used in the hml-Gal4 construct, it was important to confirm that the observed rescue was caused by tsp68C overexpression. With this purpose in mind, we designed two hml-Gal4 constructs in which the tsp68C ORF was disrupted by an early termination codon (hml^stop-Gal4), or completely absent following a truncation of the hml promoter that left only 840 bp of the hml upstream sequence (hml-Gal4) (Figure 2a and c). These constructs were injected into fly embryos and transgenic lines were established. Both hml-Gal4 and hml^stop-Gal4 stocks showed the characteristic Gal4 expression pattern described for the original hml-Gal4 lines, with robust and specific expression in larval lymph glands and circulating hemocytes (Figure 2d and data not shown). Based on these findings, we conclude that most, if not all of the regulatory elements responsible for the hematopoietic-specific expression are contained within the short 840 bp fragment. Interestingly, there are subtle differences in larval hemocyte composition and morphology between the old and modified hml-Gal4 lines. Hemocytes isolated from the modified hml-Gal4 lines display a range of cell morphology that parallels that of wild-type hemocytes. In contrast, the majority of hemocytes in the original hml-Gal4 line consist predominantly of small, uniform round plasmatocytes with rare cell membrane protrusions or pseudopodia (Figure 2d). In addition, lamellocytes (a rare but detectable subset of hemocytes under normal conditions) are very rarely detectable in hml-Gal4 stocks (data not shown).

An hml^Stop-Gal4 insertion was recombined with the ytr^P22 mutant chromosome. In contrast to what we had observed with the original hml-Gal4, ytr recombinants, none of the modified hml^Stop-Gal4, ytr^P22 recombinants were able to rescue the hematopoietic defects in the mutant homozygous state (Figure 3). There was complete concordance in hemocyte morphology and composition in the hemolymph of hml^Stop-Gal4, ytr^P22 and the original ytr^P22 mutant. Lymph glands were also equally enlarged and populated with abnormal hemocytes (data not shown). However, we completely restored a normal hematopoietic profile when we provided a wild-type copy of the ytr gene in hml^Stop-Gal4, ytr^P22 homozygous larvae, confirming both the specificity of our modified hml^Stop-Gal4 line and the cell autonomy nature of the defects in the hematopoietic tissue of ytr mutants (Figure 3). These data strongly suggest that overexpression of tetraspanin tsp68C in ytr mutant hemocytes restores their normal differentiation and morphology.

Figure 3.

Modified hml^Stop-Gal4 line fails to rescue the proliferation and differentiation defects in ytr mutant hemocytes. (a) Analysis of hemocyte morphology and (b) hemocyte counts are shown for each genetic background. They were performed as described in Materials and methods. To ensure reproducibility, multiple (10) independent samples of each genotype were analysed

Full figure and legend (127K)

Tsp68C is a member of the four transmembrane protein tetraspanin family and localizes to the perinuclear and trans Golgi compartments of cells

RT–PCR analysis showed that two forms of tsp68C mRNAs are expressed in wild-type hemocytes (Figure 2b). Sequencing of both cDNAs revealed that they consist of a spliced (short) and unspliced (long) forms that encode two proteins, Tsp68C-A and Tsp68C-B, respectively, that differ slightly in their N-termini (Figure 2b). Based on protein homology, Tsp68C is a tetraspanin that is more closely related to the human leukocyte differentiation antigens CD151 (19% identity) and CD63 (15%) (Figure 4a). Interestingly, Tsp68C is quite distant from other fly tetraspanins (Todres et al., 2000). It only shares 20–23% identity with its closest members, Tsp42Ed, Tsp96F and Tsp5D. Furthermore, Tsp68C is unusual in having a long C-terminal intracellular tail (78 a.a.). This region contains a number of distinct short motifs that share limited homology with the ubiquitin-like protein NEDD8 and actin binding filamin-type protein (data not shown).

Figure 4.

Tsp68C is closely homologous to the human leukocyte tetraspanins CD151 and CD63. (a) Protein sequence alignment of Tsp68C with CD151 and CD63. Tetraspanin specific, highly conserved amino-acid residues are shaded in orange. (b) Localization of HA tagged Tsp68C protein (HA- Tsp68C) in cytoplasmic trans-Golgi compartments of NIH 3T3 cells. Control transfection in NIH 3T3 cells confirm localization of the HA tagged MITF transcription factor (HA-MITF) to the nucleus. In all panels, the nucleus is visualized after DAPI staining and the protein of interest is detected with fluorescein-labeled anti HA-antibody (see Materials and methods)

Full figure and legend (324K)

Many tetraspanins, including human CD151 and CD63, localize to the lysosomal and trans Golgi perinuclear compartments of cells. To determine the subcellular distribution of Tsp68C, an HA tagged version of Tsp68C was expressed in NIH 3T3 fibroblasts (see Materials and methods). Like CD63 and other intracellular mammalian tetraspanins (Hubner et al., 2002; Mantegazza et al., 2004), HA-Tsp68C is not diffusely localized throughout the cell, but is detected in a 'web-like' pattern that extends from the cell surface to the perinuclear space (Figure 4b). Control transfection in which HA-tagged MITF transcription factor was expressed, confirmed that under our experimental conditions, MITF localized to its proper nuclear compartment (King et al., 1999).

tsp68C overexpression also suppresses the effect of activated Raf in hemocytes

To gain further insight into the function of tsp68C, we extended our studies on the effect of overexpression of this gene in hemocytes by simultaneously expressing in these cells activated forms of the Raf or Hop (the Jak equivalent in flies) kinases. Hyperactivation of either the Ras/Raf/MAPK or the Jak/STAT signal transduction pathway causes overproliferation and abnormal differentiation of Drosophila hemocytes (Luo et al., 1995; Asha et al., 2003). We compared the effect on larval hemocytes when the original hml-Gal4 or our newly modified hml-Gal4 lines were used to direct expression of the constitutively active forms of the cRaf1^gof or Hop^Tum-l kinases. Stocks carrying either the UAS-cRaf1^gof (Brand and Perrimon, 1994) or UAS-hop^Tum-l (Bach et al., 2003) transgene were crossed with the original hml-Gal4 or with hml^stop-Gal4 or hml-Gal4 lines. With activated Raf, we observed a dramatic difference in hemocyte proliferation depending on the type of hml-Gal4 used (Figure 5a and b). There was a 50- to 100-fold elevation in the number of circulating larval hemocytes (up to 250 10³ cells per larva) when activated Raf was expressed through the hml^stop-Gal4 and hml-Gal4 lines (Figure 5b). In contrast, expression of activated Raf through the original hml-Gal4 resulted only in a modest elevation of hemocyte counts (up to 25–35 10³ cells/larva) (Figure 5a and b). Proliferation of larval hemocytes induced after expression of activated Ras was also significantly suppressed in crosses where hml-Gal4 was used as compared to hml^Stop-Gal4 (Figure 5c). The difference in proliferation cannot be due to differences in expression levels mediated by the various hml-Gal4 lines. As shown for crosses with UAS-cRaf1^gof, we failed to observe any dramatic difference in levels of activated Raf protein or mRNA in the hemocytes of affected larvae (Figure 5d and data not shown). Therefore, the suppression of Raf- (or Ras-) induced proliferation in larvae carrying the original hml-Gal4 construct correlates with overexpression of tsp68C. In contrast, overexpression of tsp68C did not interfere with the effect of Hop^Tum-l in larval hemocytes. In all cases, there was a similar increase in circulating hemocytes and a robust accumulation of lamellocytes, irrespective of the type of hml-Gal4 line used (Figure 5e and f). These data suggest that Tsp68C can specifically modulate the strength of signals emanating from the activation of the Ras/Raf/MAPK pathway in hemocytes but does not modulate Jak/STAT signaling in these cells.

Discussion

The large number of tetraspanin genes in each species (37 for Drosophila and 28 for humans) and their likely overlap in function has significantly hampered loss of function studies. Although redundancy is the likely culprit for the lack of detectable phenotype, it is possible that despite their strong conservation across species and their high prevalence, tetraspanins are for the most part dispensable for viability and other important functions. This view, however, is challenged by the observation that the level of expression of many human tetraspanins influences the proliferation, motility and metastatic potential of different tumors (Boucheix et al., 2001).

Like vertebrate blood cells, Drosophila hemocytes express a number of tetraspanin genes, including tsp42Ed, tsp42El, tm4sf and tsp68C (Fradkin et al., 2002; Asha et al., 2003; SA Sinenko and B Mathey-Prevot, (unpublished data). Their role in these cells was completely unknown. We came across tsp68C rather serendipitously as this gene is nested within the hemolectin promoter used in a construct carried by transgenic flies (hml-Gal4) that expressed Gal4 specifically in hemocytes (Goto et al., 2001). We established that tsp68C was specifically upregulated in the lymph gland and the circulating hemocytes of hml-Gal4 larvae. To our surprise, overexpression of tsp68C was sufficient to restore a near-normal hematopoietic profile in ytr mutants, which typically show dramatic proliferation and differentiation defects in these cells (Sinenko et al., 2004). The rescue was not due to the ectopic expression of Gal4 in these cells but to the increased levels of tsp68C. When expression of tsp68C or its gene product was abolished in two modified hml-Gal4 lines, we observed no hematopoietic rescue in ytr mutants. Although, the exact mechanism(s) of how increased levels of Tsp68C in hemocytes restores normal function in ytr mutant hemocytes will require a knowledge of the precise cellular and biochemical function of the Ytr protein, our data suggest that tsp68C acts as a tumor-suppressor gene in hemocytes in the context of a ytr mutation and provide the first example of a functional role for tetraspanins in Drosophila hematopoiesis. This finding is in agreement with the observation that the function or changes in expression of a number of mammalian tetraspanins can affect hematopoiesis and/or tumorigenesis. Tsp CD9 and CD63 have been functionally implicated in megakaryocyte differentiation and dendritic cell maturation, respectively (Boucheix and Rubinstein, 2001; Engering et al., 2003). Both TSSC6 and CD81 proteins have been described to play a role in B- and T-cell function and affect their proliferation (Miyazaki et al., 1997; Tarrant et al., 2002), and repression of CD9, CD82, or CD63 expression correlates with advanced, metastatic stages of different human tumors (Boucheix and Rubinstein, 2001).

To gain additional insight into how Tsp68C might affect proliferation in hemocytes, we examined whether its overexpression could affect signaling by oncogenic Ras or by constitutively active forms of the Raf and Hop kinases. Interestingly, when we expressed activated Raf (cRaf1^gof) or Ras (Ras^85D.V12) in hemocytes using the original hml-Gal4 line, we only observed a mild increase in proliferation. This is in contrast to the massive overproliferation caused after expression of these proteins with modified hml-Gal4 lines that no longer coexpressed tsp68C. Interestingly, increased expression of tsp68C failed to suppress the effect of Hop^Tum-l in hemocytes. Thus, the net increase in tsp68C expression mediated by the hml-Gal4 line was sufficient to suppress abnormal proliferation of hemocytes triggered by the activated Raf and Ras proteins but did not protect against the effect of Hop^Tum-l.

The specific effect of a tetraspanin on the Ras/Raf/MAPK pathway is not unprecedented. Many human tetraspanins have been implicated as facilitators in the assembly of effector proteins or in the relay of signals from this pathway. Indeed, physical association between CD151, CD63, CD81, CD9 and various integrins, the c-Kit receptor, PKC or phosphoinositide 4-kinase has been documented in a cell-specific manner and proposed to modulate either the TGF- or Ras/Raf/MAPK pathway activity (Shi et al., 2000; Zhang et al., 2001; Anzai et al., 2002; Sawada et al., 2003; Carloni et al., 2004). Similarly, we propose that Tsp68C might achieve its regulation by targeting or trapping specific effector proteins to specialized compartments, where they are unable to participate in the propagation of a signal. The cellular localization of Tsp68C is consistent with this view as it is similar to that of other tetraspanins which have been implicated in intracellular trafficking (Berditchevski, 2001; Mantegazza et al., 2004).

Although increased level of Tsp68C expression can specifically suppress constitutive activation of the Ras/Raf/MAPK pathway in hemocyte cells, we did not observe any significant deficiency in hemocyte numbers in the original hml-Gal4 line. We hypothesize that under normal conditions, hemocytes do not entirely rely on activation of this pathway, and that other pathways compensate for reduced Ras/Raf/MAPK signaling. In that respect, proliferation of hemocytes in larvae is known to be affected by the Toll or Jak/STAT pathways (Sorrentino et al., 2004) and is influenced by the release of ecdysone hormone during development (Sorrentino et al., 2002). However, hemocytes in the original hml-Gal4 lines are not completely normal. The presence of a highly uniform population of plasmatocytes and the virtual absence of lamellocytes are in sharp contrast with the cellular plasticity and diversity exhibited by wild-type hemocytes, and suggest that overexpression of tsp68C modifies normal hemocyte morphology and perhaps restricts their differentiation toward certain lineages.

In summary, our results provide evidence that increased level of tsp68C expression can overcome a range of signaling defects in hemocytes, and in particular can specifically suppress constitutive activation of the Ras/Raf/MAPK pathway in these cells. In addition, our experiments validate the use of gain of function approaches in unraveling the function of tetraspanins and open up the possibility, in conjunction with the UAS/Gal4 system, to carry out structure/function studies of Tsp68C to delineate critical regions in this protein.

Materials and methods

Drosophila stocks and genetics

The ytr mutant alleles and UAS-ytr constructs have been recently described (Sinenko et al., 2004). All other stocks, including hml-Gal4 lines (Goto et al., 2001), UAS-cRaf1^gof (Brand and Perrimon, 1994) and UAS-Ras^85D.V12 (Asha et al., 2003) were obtained from the Bloomington Stock Center. UAS-hop^Tum-l (Bach et al., 2003) flies were kindly provided by Erika Bach and Herve Agaisse. Oregon-R and w flies were used as wild-type controls. Flies were maintained on standard Drosophila cornmeal/sucrose/yeast medium at 20 or 25°C as necessary.

Design of new hml-Gal4 lines to silence or delete the tsp68C gene

KpnI and BamHI sites were added to the original forward and reverse primers described by Goto et al. (2001) to facilitate cloning into the pGAT vector. A TAG stop codon was introduced in the coding sequence of the second exon of tsp68C by PCR, using a reverse (5'-CTCAAGCTTCTATGATACTACCAGCAGCAAG-3') primer and a forward (5'-TAGAAGCTTTTGCTGCTGGTAGTATCAGGTCTA-3') primer containing an HindIII recognition site. The KpnI/NotI fragment corresponding to the mutated or truncated hml promoter was inserted upstream of the Gal4 ORF in the CaSpeR4 vector. Transgenic lines were generated as previously described (Callus and Mathey-Prevot, 2002).

tsp68C transcripts were quantified following the one-step RT–PCR reaction kit (Invitrogen). For each reaction, the tsp68C forward primer (5'-GGCCTGCTGTTTTAATTACAAA-3') and reverse primer (5'CAACTTAGCGGCACCGGCCAATT3') were used in conjunction with 1 g of total RNA. RNA was isolated from larvae or from hemocytes with TRIzol^R (Invitrogen) and DNA-free RNA kit (ZYMO Research). The PCR reaction was carried out for 24 cycles, with each cycle consisting of denaturation (94°C for 30 s), annealing (56°C for 30 s), and extension (72°C for 1 min).

Hemocyte preparations and counts

Hemocytes from a single larva of each genotype were collected in 50 l of Ringer solution in a single well of a 10-well glass slide (PGC Scientifics). Cells were let to adhere before fixation with a 2% formaldehyde solution (Sigma) and final preparations were mounted in 80% glycerol in PBS. Multiple (10) independent samples for each genotype were analysed to ensure reproducibility. To estimate expression of human cRaf1^gof in hemocytes, fixed cells were treated with rabbit anti-cRaf1-specific antibodies (C12, Upstate), followed by staining with Alexa⁵⁶⁸ labeled anti-rabbit antibodies (Molecular Probes). Cell images were taken under different magnifications with a light (Nomarski optics) and fluorescent microscope (Zeiss Optical). To determine total hemocyte numbers in different mutants, the hemocytes from five larvae of each genotype were collected in 50 l of Ringer solution. The resuspended cells were counted using a hemocytometer and counts were used to determine the number of cells per single larva.

Cell culture and transfection

NIH 3T3 cells were cultured in RPMI supplemented with 10% FCS. Transfections were carried out with the FUGENE-6 transfection reagent (Roche) as was described (Sinenko et al., 2004). Briefly, subconfluent cultures of cells grown on glass slides (Lab-Tek II Chamber slide system, Nalge Nunc, Int.) were transfected with 2.5 l of the FUGENE-6 reagent containing either 0.5–1 g of HA-Tsp68C- and HA-MITF- pcDNA3.1 plasmids in 200 l serum-free RPMI. RPMI containing 10% FCS medium was added 1 h later. At 24 h after transfection, cells were fixed with 3% formaldehyde in PBS for 20 min and incubated in 0.1% Tween-20, 0.3% BSA in PBS. Fluorescein labeled HA-specific monoclonal antibodies (3F10, Roche) were added to the fixed and permeabilized cells for 2 h, and DNA was counterstained with DAPI. Cells were then examined by fluorescence microscopy (Zeiss Optical).

References

1. Anzai N, Lee Y, Youn BS, Fukuda S, Kim YJ, Mantel C, Akashi M and Broxmeyer HE. (2002). Blood, 99, 4413–4421. | Article | PubMed |
2. Asha H, Nagy I, Kovacs G, Stetson D, Ando I and Dearolf CR. (2003). Genetics, 163, 203–215. | PubMed | ISI | ChemPort |
3. Bach E, Vincent S, Zeidler M and Perrimon N. (2003). Genetics, 165, 1149–1166. | PubMed | ISI | ChemPort |
4. Berditchevski F. (2001). J. Cell Sci., 114 (Part 23), 4143–4151. | PubMed | ISI | ChemPort |
5. Boucheix C, Duc G, Jasmin C and Rubinstein E. (2001). Expert Rev. Mol. Med., 2001, 1–17. | Article | PubMed | ChemPort |
6. Boucheix C and Rubinstein E. (2001). Cell Mol. Life Sci., 58, 1189–1205. | Article | PubMed | ISI | ChemPort |
7. Brand A and Perrimon N. (1993). Development, 118, 401–415. | PubMed | ISI | ChemPort |
8. Brand A and Perrimon N. (1994). Genes Dev., 8, 629–639. | PubMed | ISI | ChemPort |
9. Callus BA and Mathey-Prevot B. (2002). Oncogene, 21, 4812–4821. | Article | PubMed | ChemPort |
10. Carloni V, Mazzocca A and Ravichandran K. (2004). Oncogene, 23, 1566–1574. | Article | PubMed | ISI | ChemPort |
11. Engering A, Kuhn L, Fluitsma D, Hoefsmit E and Pieters J. (2003). Eur. J. Biochem., 270, 2412–2420. | Article | PubMed | ISI | ChemPort |
12. Fradkin LG, Kamphorst JT, DiAntonio A, Goodman CS and Noordermeer JN. (2002). Proc. Natl. Acad. Sci. USA, 99, 13663–13668. | Article | PubMed | ChemPort |
13. Goto A, Kumagai T, Kumagai C, Hirose J, Narita H, Mori H, Kadowaki T, Beck K and Kitagawa Y. (2001). Biochem. J., 359, 99–108. | Article | PubMed | ISI | ChemPort |
14. Hemler ME. (2001). J. Cell Biol., 155, 1103–1107. | Article | PubMed | ISI | ChemPort |
15. Hubner K, Windoffer R, Hutter H and Leube R. (2002). Int. Rev. Cytol., 214, 103–159. | Article | PubMed |
16. King R, Weilbaecher KN, McGill G, Cooley E, Mihm M and Fisher DE. (1999). Am. J. Pathol., 155, 731–738. | PubMed | ISI | ChemPort |
17. Kopczynski CC, Davis GV and Goodman CS. (1996). Science, 271, 1867–1870. | Article | PubMed | ISI | ChemPort |
18. Luo H, Hanratty WP and Dearolf CR. (1995). EMBO J., 14, 1412–1420. | PubMed | ISI | ChemPort |
19. Mantegazza A, Barrio M, Moutel S, Bover L, Weck M, Brossart P, Teillaud J and Mordoh J. (2004). Blood, 104, 1183–1190. | Article | PubMed | ISI | ChemPort |
20. Miyazaki T, Muller U and Campbell K. (1997). EMBO J., 16, 4217–4225. | Article | PubMed | ISI | ChemPort |
21. Sawada S, Yoshimoto M, Odintsova E, Hotchin N and Berditchevski F. (2003). J. Biol. Chem., 278, 26323–26326. | Article | PubMed | ISI | ChemPort |
22. Scherberich A, Giannone G, Perennou E, Takeda K, Boucheix C, Rubinstein E, Lanza F and Beretz A. (2002). Thromb. Haemost., 87, 1043–1050. | PubMed |
23. Shi W, Fan H, Shum L and Derynck R. (2000). J. Cell Biol., 148, 591–602. | Article | PubMed | ISI | ChemPort |
24. Sinenko SA, Kim EK, Wynn R, Manfruelli P, Ando I, Wharton KA, Perrimon N and Mathey-Prevot B. (2004). Dev. Biol., 273, 48–62. | Article | PubMed | ISI | ChemPort |
25. Sorrentino RP, Carton Y and Govind S. (2002). Dev. Biol., 243, 65–80. | Article | PubMed |
26. Sorrentino RP, Melk JP and Govind S. (2004). Genetics, 166, 1343–1356. | Article | PubMed | ISI | ChemPort |
27. Tarrant J, Robb L, van Spriel A and Wright M. (2003). Trends Immunol., 24, 610–617. | Article | PubMed | ISI | ChemPort |
28. Tarrant JM, Groom J, Metcalf D, Li R, Borobokas B, Wright MD, Tarlinton D and Robb L. (2002). Mol. Cell Biol., 22, 5006–5018. | Article | PubMed | ISI | ChemPort |
29. Todres E, Nardi J and Robertson H. (2000). Insect Mol. Biol., 9, 581–590. | Article | PubMed | ISI | ChemPort |
30. Xu H, Lee S, Suzuki E, Dugan K, Stoddard A, Li H, Chodosh L and Montell C. (2004). EMBO J., 23, 811–822. | Article | PubMed | ISI | ChemPort |
31. Yunta M and Lazo P. (2003). Cell Signal., 15, 559–564. | Article | PubMed | ISI | ChemPort |
32. Zhang XA, Bontrager AL and Hemler ME. (2001). J. Biol. Chem., 276, 25005–25013. | Article | PubMed | ISI | ChemPort |

Acknowledgements

We thank Melissa Sackal, Pascal Manfruelli and members of the Perrimon laboratory for helpful discussion on this project and sharing fly stocks and reagents. We are grateful to Dr Norbert Perrimon, Dr Martin Hemler and Dr Utpal Banerjee for discussion and helpful comments about these studies. We thank for the expertise of C Villalta in performing embryo injections. We acknowledge the Bloomington Stock Center for providing us with numerous fly stocks. This work was supported in part by Grant R01 HL62434 and the Claudia Adams Barr Program in Cancer Research (BM-P) and David Abraham Fellowship (SAS)