DEV02091 Supplementary Material

Files in this Data Supplement:

• Movie 1 -

  Movie 1. Time lapse of the heart remodelling analysed on a NP1029-Gal4; tinCD4-Gal4; UAS-mcd8-GFP pupae between 24 and 54 hours APF. NP 1029-Gal4 driver is expressed in the larval cardiac tube but fades during metamorphosis, while the tinCD4-Gal4 driver, which is not expressed during larval life, is activated during cardiac tube remodelling (see Figs S1 and S2 for a more complete description of the expression patterns of drivers). Combining both drivers allowed cardiac tube remodelling to be followed throughout metamorphosis and shows that adult heart formation is a progressive process that results from remodelling of the larval cardiac tube myocytes. The elapsed time after puparium formation is indicated at the bottom right. Up to 30 hours APF, the larval heart is visible by 1029-Gal4 expression. At 30 hours APF, the larval aorta comes into focus due to its dorsal repositioning. Then GFP expression in the larval heart progressively fades, owing to both NP1029-Gal4 expression fading and to the elimination of the posterior-most myocytes by programmed cell death. Meanwhile, the tincD4-Gal4 driver is activated in all myocytes anterior to A5 svp myocytes, allowing the subsequent steps of adult heart formation to be followed. Arrowhead indicates a developing ostium. The cardiac tube is enlarged, while the larval aorta is remodelled into adult heart.

• Movie 2 -

  Movie 2. Time lapse of the heart remodelling analysed on a NP1029-Gal4; tinCD4-Gal4; UAS-mcd8-GFP, UAS-nls-dsRed pupae between 27 to 64 hours APF. The drivers control both UAS-mcd8-GFP and UAS-nls-dsRed, allowing following both the general morphology of the cardiac tube and the myocyte nuclei. The time lapse starts just after larval aorta dorsal re-localisation and focuses on segment A3/A4 junction. From top to bottom: dynamics of nls-dsRed expression, merge and dynamics of mcd8-GFP expression. Top: the four pairs of A3 Tin myocytes are indicated all along the film (asterisk), unambiguously indicating that adult myocytes are identical to larval myocytes, which are remodelled without cell division. NP1029-Gal4 is not expressed in svp cells, thus it was not possible to follow their fate during remodelling. In this particular example, tincD4-Gal4 activation is not visible in svp myocytes. Bottom: mcd8-GFP, while weakly expressed, allows the general morphology of the cardiac tube to be followed during its remodelling. Enlargement of the tube is clearly visible, as well as enrichment of GFP in the developing valves (open arrowheads).

• Movie 3 -

  Movie 3. General view of a beating larval cardiac tube. Larval cardiac tube expressing both nuclear and cytoplasmic dsRed under the NP 1029-Gal4 control. This driver is expressed in all Tin myocytes except the cardiovascular valve. Recording was performed on dissected third instar larvae. The heart (segments A5-A7) has a stronger and more rhythmic activity than the aorta (segments A1 to A4).

• Movie 4 -

  Movie 4. Detail (segments A3/A5) of a beating larval cardiac tube. Larval cardiac tube expressing both nuclear and cytoplasmic dsRed under the NP 1029-Gal4 control. Recording was performed on dissected third instar larvae. Detail of segments A3-A5, emphasizing the functional differences between the larval aorta and the heart. In particular, while the A5 svp cells form functional ostiae that open and close in phase with heart beat (arrowhead), the svp cells in segments A3 and A4 are unable to open (arrow). Open arrowhead indicate the valve of segment A4, which do not express the NP-1029 driver.

• Movie 5 -

  Movie 5. General view of a beating adult cardiac tube. Dissected newly eclosed adult expressing membrane bound GFP (mcd8-GFP) under both 24B-Gal4 and tinCD4-Gal4 control. At adulthood, 24B is expressed in all myocytes and in tinCD4-Gal4 in myocytes anterior to A5 with marked accumulation in valves and ostias. General view of the adult cardiac tube, showing that segments A1-A4, which constitute the adult heart, have acquired strong and rhythmic activity during remodelling (compare with the same segments in Movie 3).

• Movie 6 -

  Movie 6. Detail (segments A3/A4) of a beating adult cardiac tube. Membrane bound GFP driven by tinCD4-Gal4 analysed on an intact newly eclosed adult. GFP accumulates predominantly in valves (open arrowhead) and the ostia (arrowhead), which open and close in rhythm with heart beating, in marked contrast with their activity during larval life (compare with Movie 4). To allow the best visualisation of heart movements the movie has been slowed down.

• Movie 7 -

  Movie 7. Detail (segments A4/A5) of a beating adult cardiac tube. Membrane-bound GFP driven by 24B-Gal4 analysed on dissected newly eclosed adult. Top: a focus on A4 segment illustrates the strong beating activity of the adult heart (arrowhead point to A5 ostia). Bottom: same preparation, but focused on terminal chamber (arrow). Compared with anterior segments, segment A5 has only weak beating activity at adulthood. Most of the terminal chamber movements are a consequence of the anterior heart beating.

• Supplemental Figure 1 -

  Fig. S1. Timetable of cardiac tube remodelling. Expression pattern of a membrane-bound GFP driven by NP1029-Gal4 (A,B) or by tinC?4-Gal4 (C,D) at the indicated times APF. In all panels, anterior is leftwards, dorsal is upwards and arrowheads indicate the anterior limit of segment A5. (A) At 24 hours APF, the cardiac tube retains larval morphology and function with the contracting heart lying in segments A5 to A7 (arrow); the posterior aorta, lying in segments A1 to A4, is detached from the dorsal cuticle and is out of focus. (B) Cardiac tube remodelling coincides with the last ecdysone peak at 30 hours APF (Riddiford, 1993). Heart beating stops between 27 and 30 hours APF and the posterior aorta undergoes a dorsal migration, to contact the dorsal epidermis (indicated by an arrow). Histolysis of larval heart segments A6-A7 also starts at 30 hours APF (see Molina and Cripps, 2001). NP1029-Gal4 expression fades after 30 hours APF and is no longer detected at 48 hours APF (data not shown), nor at adulthood. (C,D) tinC?4-Gal4 expression pattern during metamorphosis. Once the cardiac tube has completed its dorsal repositioning, the developing adult heart progressively enlarges from 36 hours to 60 hours APF. tinC?4-Gal4 is expressed in all the embryonic myocytes of the cardiac tube, including svp-expressing myocytes (Lo and Frasch, 2001), but it is not expressed in the larva (not shown). Expression is reinitiated from 30 hours APF onwards in all myocytes anterior to A5 (C), allowing a clear visualisation of the developing adult heart, including enlargement of the tube and valves, and ostiae morphogenesis (open arrow head and asterisk in D, respectively). At 60 hours APF, the first signs of adult cardiac activity are detectable while ostiae become functional at about 72 hours APF (not shown). At 96 hours APF, steady heart beating is observed (see Movies 5-7).

• Supplemental Figure 2 -

  Fig. S2. Origin and formation of the adult cardiac tube. (A-C) Cell-tracing experiment. (A) NP[1029]-Gal4; UAS nls-GFP larval cardiac tube stained for GFP (green), Mef2 (red) and F-actin (blue). During larval stages, the driver is expressed in all Tin-expressing myocytes, except in the second pair in segment A4 (asterisk). The svp cells are Mef2 positive, do not express the driver and are recognized by the small size of their nuclei (arrowheads). (B) NP[1029]-Gal4; UAS nls-GFP adult cardiac tube stained for GFP (green), Mef2 (red) and F-actin (blue). At adulthood, the NP[1029]-Gal4 driver is no longer expressed in the cardiac tube. (C) NP[1029]-Gal4; UAS nls-bGal adult cardiac tube with X-gal staining. Owing to its stability, b-gal driven in the cardiac tube by the NP[1029]-Gal4 driver is still detectable at adulthood in the Tin-myocytes, except in the second pair in segment A4 (asterisk) and in svp-expressing cells (arrowheads), allowing an identity to be attributed to each adult myocyte of the cardiac tube with respect to its larval origin (see schematic in Fig. 1C). The adult heart is formed by myocytes that constituted the larval posterior aorta (segments A1 to A4) and by segment A5 myocytes, which formed part of the larval heart and that constitute the posterior tip (terminal chamber) of the adult heart. (D-G) Cellular aspects of heart remodelling. Unless otherwise stated, arrowheads indicate svp-expressing cells, arrows indicate Tin myocytes and open arrowheads indicate the valves. (D,E) Part of a larval (D) and adult (E) cardiac tube that express a membrane-bound GFP (mcd8>GFP) under 24B-GAL4 (D) or tinCD4-Gal4 (E) control, stained for Tin (red) and visualised at the same magnification. Each segment is composed of four Tin-expressing and two Tin-negative myocytes pairs (svp cells), the latter being visualised with DAPI (red arrowhead, insert) in D [lg/pc: DAPI staining also reveals lymph glands (lg) and pericardiac (pc) cells associated with the larval cardiac tube]. (F,G) Phalloidin staining of A3/A4 segments of a larval cardiac tube (F), and its corresponding region at adulthood (G) at the same magnification (svp cells: green dotted line, they are opened in G; Tin-myocytes: blue dotted line; valves: yellow dotted line). During remodelling, svp-expressing myocytes develop functional ostiae: they increase in size, lose their adhesion to the adjacent svp-expressing cell (arrowhead, compare D,F with E,G) and, while they are non-functional during larval stage (see Movies 3 and 4), they gain the ability to open and close, thereby becoming able to control the entry of haemolymph (see Movies 5 and 6). Likewise, Tin myocytes in segments A1-A4 also increase in size, develop a denser myofibril network (compare D,F and E,G) and, although they are non-contracting automatically in larvae (see Movies 3 and 4), they acquire the automatic rhythmic beating of typical heart myocytes (see Movies 5 and 6). In addition, the second pair of Tin myocytes in segments A2, A3 and A4 develops valves, characterized by a very dense F-actin web (open arrowheads in E, G). Larval myocytes of the posterior aorta thus undergo extensive morphological and functional changes while they are remodelled to form the adult heart. (H,I) Phalloidin staining of A5 segment in the larval cardiac tube (H) and in the adult (I) at same magnification. (H) In the larva, A5 myocytes contain dense transversal (red arrow) myofibrils. (I) At adulthood, A5 myocytes which constitute the terminal chamber, are thinner and their myofibrils are orientated longitudinally (red arrow), indicating a change in cell polarity during remodelling. These cellular shape modifications are linked to functional changes, and while actively beating during larval life (see Movies 3 and 4), A5 myocytes display poor contracting activity at adulthood (see Movie 7). Segment A5 thus loses its heart characteristics during metamorphosis to differentiate into new types of myocytes constituting the terminal chamber with no equivalent in the larva. (J-M) Cardiac A5 segments gain specific innervations at adulthood. (J) Ventral view of an adult cardiac tube stained for polymerised actin (phalloidin, white) and nerve terminations (22C10, red). Longitudinal fibres along A1 to A4 segments correspond to ventral imaginal muscles (red arrow) (described by Molina and Cripps, 2001). 22C10 staining reveals nerve terminations along the whole cardiac tube, the most robust being localized in A5 (white open arrow). (K-M) Details of an adult cardiac tube stained for phalloidin (white), 22C10 (red) and synaptotagmin (green). (K,L) Two different focal planes of the same part of the cardiac tube (A3 segment). In segments anterior to A5, nerve terminations make synaptic contacts (revealed by synaptotagmin) targeted to imaginal longitudinal muscles (red arrow in K) but not to heart myocytes (L). (M) No imaginal longitudinal muscles develop beneath A5 (Molina and Cripps, 2001), and in this segment nerve terminations directly contact the myocytes (white open arrows). These specific innervations of A5 myocytes are gained during metamorphosis (Dulcis and Levine, 2003; Rizki, 1978) (our own observations). This anteroposterior patterning of adult heart innervations suggests that A5 nerve terminations may play specific physiological function and may well be involved in the regulation of the anterograde beating. (K-M) Confocal images. Scale bars: 50 mm.