A misexpression screen to identify regulators of Drosophila larval hemocyte development

Abstract

In Drosophila, defense against foreign pathogens is mediated by an effective innate immune system, the cellular arm of which is composed of circulating hemocytes that engulf bacteria and encapsulate larger foreign particles. Three hemocyte types occur: plasmatocytes, crystal cells, and lamellocytes. The most abundant larval hemocyte type is the plasmatocyte, which is responsible for phagocytosis and is present either in circulation or in adherent sessile domains under the larval cuticle. The mechanisms controlling differentiation of plasmatocytes and their migration toward these sessile compartments are unclear. To address these questions we have conducted a misexpression screen using the plasmatocyte-expressed GAL4 driver Peroxidasin-GAL4 (Pxn-GAL4) and existing enhancer-promoter (EP) and EP yellow (EY) transposon libraries to systematically misexpress approximately 20% of Drosophila genes in larval hemocytes. The Pxn-GAL4 strain also contains a UAS-GFP reporter enabling hemocyte phenotypes to be visualized in the semitransparent larvae. Among 3412 insertions screened we uncovered 101 candidate hemocyte regulators. Some of these are known to control hemocyte development, but the majority either have no characterized function or are proteins of known function not previously implicated in hemocyte development. We have further analyzed three candidate genes for changes in hemocyte morphology, cell-cell adhesion properties, phagocytosis activity, and melanotic tumor formation.

Figure 1.—

Hemocyte expression of the Pxn-GAL4 driver. (A) Circulating and (B) sessile hemocytes were isolated from wandering-stage Pxn-GAL4, UAS-GFP larvae and stained with antibodies against GFP (green) and the pan-hemocyte marker Hemese (red). (C) Pxn-GAL4 directed expression of GFP (green) overlaps expression of the crystal cell marker lozenge-lacZ (lz-lacZ) (red). (D) Lamellocyte overproduction was triggered by introducing one copy of the hop^Tum-l mutation into the Pxn-GAL4, UAS-GFP background. Lamellocytes (arrowheads, revealed by MAb L1b staining in red) do not express GFP (green). GFP-expressing plasmatocytes (asterisks) are not MAb L1b positive. Bar, 20 μm.

Figure 2.—

Temporal profile of hemocyte accumulation. GFP-expressing hemocytes were visualized in Pxn-GAL4, UAS-GFP first, second, and third instar larvae. (A) In first instar larvae a sessile population of hemocytes forms at the posterior of the larva (arrowhead). (B) By the second instar larval stage, this posterior accumulation (arrowhead) is accompanied by the formation of distinct segmentally repeated dorsal patches or compartments (asterisks). (C) Third instar larvae show an increased number of hemocytes forming distinct dorsal sessile patches (asterisks) and accumulations at the tail region (arrowhead).

Figure 3.—

Schematic of the screen. The Pxn-GAL4, UAS-GFP driver line was crossed to a set of EP and EY lines. The GFP blood expression pattern was recorded and the lines that show deviations from the parental pattern were retained (positives) and retested. For lines that pass the retest, EP/EY lines that contain insertions in the same gene or in close proximity were scored again for identical blood phenotypes.

Figure 4.—

Observed hemocyte phenotype classes. (A) Hemocyte distribution in wild-type third instar larvae. Principal overexpression phenotypes were (B) disruption of dorsal sessile hemocyte compartments; (C) increases in lymph gland size; (D) increases in hemocyte number; (E) relocalization of hemocytes to the dorsal vessel; and (F) spreading of hemocytes through the cuticle. Hemocytes are indicated as circles, lymph glands (lg) are shaded in the anterior of the larva, the dorsal vessel (dv) runs the length of the animal. Thoracic and abdominal segments are indicated (T1–A8).

Figure 5.—

Selected overexpression phenotypes and gene ontology classification of candidate genes. (A) Hemocyte distribution in control Pxn-GAL4, UAS-GFP third instar larvae. (A′) Detail of the posterior of the same larva showing the distinct dorsal sessile compartments. (B) Pxn-GAL4 directed ectopic expression of Kr changes hemocyte distribution and morphology. (B′) Large flattened cells are observed. (C) Ectopic expression of nej results in generally weaker GFP expression in most hemocytes in both circulating and sessile populations apart from a few highly expressing cells. (D) Ectopic expression of CG32813 results in inappropriate hemocyte targeting along the posterior of the dorsal vessel (enlarged detail shown in D′). GO classification of the 101 identified candidate genes according to (E) biological process, and (F) molecular function. GO annotations were obtained from FlyBase.

Figure 6.—

Validation of selected EP/EY insertions. Expression of genes flanking EP/EY insertions identified in the screen was determined by RT–PCR. Insertions identified in the screen are indicated in boldface type while those that failed to give a hemocyte phenotype are shaded. (A) Two EP insertions in the nejire (nej) locus, EP1149 and EP1179 were selected as positives in the screen. RT–PCR on isolated hemocytes reveals that both EP1149 and EP1179 drive significant overexpression of nej but not of the flanking genes, buttonhead (btd) and CG15321. The EP insertions EP950 and EP1410 (shaded), which are inserted in an opposite orientation to EP1149 and EP1179 and are not predicted to overexpress nej, do not give a hemocyte overexpression phenotype. (B) RT–PCR confirms that the EY11357 insertion drives overexpression of Kr when crossed to Pxn-GAL4. No other predicted genes occur within 10 kb upstream or downstream of EY11357. (C) Two EY insertions in the CG32813 locus, EY07727 and EY14694, were identified in the screen. RT–PCR indicates that EY07727 drives overexpression of CG32813 when crossed to Pxn-GAL4. CG11448, which flanks CG32813, shows slight increase in expression. However, the EY06476 insertion, which is predicted to drive overexpression of CG11448, does not give a hemocyte overexpression phenotype. In A–C, He and rp49 are used as loading controls and to assess mRNA purity.

Figure 7.—

Hemocyte phenotypes following Pxn-GAL4 mediated overexpression of Kr, nej, and CG32813. (A–D) Hemocyte morphology was revealed using the pan-hemocyte marker anti-Hemese (MAb H2). In the parental Pxn-GAL4 driver a uniform population of round plasmatocytes is detected. Overexpression of nej, Kr, or CG32813 results in the appearance of lamellocytes (arrowheads) and some hemocytes with irregular profiles, presumably activated hemocytes. (E–H) Antibody staining using the lamellocyte marker MAb L1b, reveals increases in lamellocyte number following Pxn-GAL4-mediated overexpression of nej, Kr, and CG32813. (I–L) Filamentous actin (F-actin) was revealed using rhodamine-phalloidin. (I) Weak F-actin staining is detected in plasmatocytes of the Pxn-GAL4 driver line. Strong F-actin staining is detected in large hemocytes in (J) Kr-, (K) nej-, and (L) CG32813-overexpressing hemocytes, suggesting that these contain an actively polymerizing cytoskeleton network. (M–P) Antibody staining using antibodies against the Drosophila β-integrin Myospheroid (Mys) shows weak expression of Mys in control hemocytes (M), but strong expression in (N) Kr, (O) nej, and (P) CG32813 overexpressing hemocytes. Upregulation of Mys was most apparent in hemocytes >15 μm. In A–P DNA (visualized using DAPI) is shown in white and antibody- and phalloidin-staining in green. Bar, 20 μm.

Figure 8.—

Phagocytic properties of hemocytes. Hemocytes were isolated from wandering-stage third instar larvae 40 min after injection of India ink particles and visualized using phase contrast microscopy. (A) Small particles of India ink were detected in lysosomes (arrowhead) of hemocytes of the parental Pxn-GAL4, UAS-GFP driver indicating that phagocytosis was normal. (B) Larger hemocytes present after Kr overexpression did not engulf India ink particles but rather adhered to large particles of India ink (arrowheads). (C) Hemocytes overexpressing nej were still capable of engulfing India ink particles, detected in lysosomes (arrowhead). Bar, 20 μm.