Heterozygosity of ribosomal protein genes causes a variety of developmental abnormalities in humans, which are collectively known as ribosomopathies, yet the underlying mechanisms remain elusive. Here, we analyzed Drosophila Minute (M)/+ mutants, a group of mutants heterozygous for ribosomal protein genes that exhibit a characteristic thin-bristle phenotype. We found that, although M/+ flies develop essentially normal wings, simultaneous deletion of one copy of the Hippo pathway effector yki resulted in severe wing growth defects. These defects were caused by JNK-mediated cell death in the wing pouch via Eiger/TNF signaling. The JNK activation in M/+, yki/+ wing discs required the caspase Dronc, which is normally blocked by DIAP1. Notably, heterozygosity of yki reduced DIAP1 expression in the wing pouch, leading to elevation of Dronc activity. Dronc and JNK formed a positive-feedback loop that amplifies Dronc activation, leading to apoptosis. Our observations suggest a mechanism of robust tissue growth whereby tissues with reduced ribosomal protein prevent ectopic apoptosis via Yki activity.

Ribosomes, composed of ribosomal RNAs and ribosomal proteins, serve as a location for protein synthesis and thus are central for life. In humans, heterozygosity of ribosomal protein genes can cause developmental abnormalities. For instance, mutations in a variety of ribosomal protein genes, including RPS17 (Cmejla et al., 2007), RPS26 (Doherty et al., 2010) and RPS27 (Wang et al., 2015) have been identified in hematopoietic or developmental diseases, such as Diamond–Blackfan anemia, isolated congenital asplenia, and 5q− syndrome (Aspesi and Ellis, 2019; Farley-Barnes et al., 2019). However, the mechanisms of how reduced ribosomal protein gene causes such defects remain elusive.

In Drosophila, a series of heterozygous mutants for ribosomal protein genes shows a thinner-bristle phenotype with a developmental delay (Brehme, 1939; Kongsuwan et al., 1985; Marygold et al., 2007), and thus are called Minute (M) mutants. Importantly, Minute heterozygous (M/+) animals develop into essentially normal flies without any morphological defects, except for the thinner bristles, suggesting that M/+ animals exert some mechanism to overcome developmental perturbations caused by reduced ribosomal proteins. This led us to use M/+ flies as a model system in which to study the mechanisms of how reduced ribosomal protein genes cause developmental defects and how animals evade morphogenetic abnormalities under such condition.

During animal development, tissue growth is regulated by the tumor-suppressor Hippo pathway, an evolutionarily conserved kinase cascade (Misra and Irvine, 2018). In Drosophila, the Ste-20 family serine/threonine kinase Hippo phosphorylates its target Warts, which further phosphorylates a transcriptional co-activator, Yorkie (Yki), thus inhibiting translocation of Yki from the cytoplasm to the nucleus. Upon inhibition of the Hippo pathway, Yki translocates to the nucleus and binds to its target DNA sequence via a DNA-binding partner, such as Scalloped, thus inducing expression of target genes, such as Myc, Cyclin E and Diap1. Whereas upregulation of Myc or Cyclin E promotes cell proliferation, elevation of DIAP1, an E3 ubiquitin ligase, inhibits apoptosis by inactivating the Drosophila caspases Dronc, Drice and Dcp-1 (Bergmann, 2010; Ditzel et al., 2008; Lee et al., 2011). Thus, activation of Yki plays a crucial role in promoting tissue growth by promoting cell proliferation and preventing apoptosis during Drosophila development.

In this study, we found that Drosophila M/+ mutants have significant defects in wing development when one copy of the yki gene is simultaneously deleted. Mechanistically, M/+ wing discs potentially activate JNK signaling, which can induce cell death via a JNK-Dronc amplification loop, thus causing developmental defects, but this is normally prevented by Yki-mediated DIAP1 expression. Our observations propose a mechanism of robust tissue growth via Yki activity that prevents ectopic apoptosis.

Yki is required for normal wing development in M/+ flies

To study the mechanism by which ribosomal protein mutants achieve robust morphogenesis during development, we conducted a genetic screen in Drosophila for dominant mutations that induce developmental defects in animals heterozygous for the ribosomal protein gene RpS3, one of the M genes. We screened 322 mutant strains and found that 15 chromosomal deficiency lines and two mutant lines resulted in notched-wing phenotypes when combined with RpS3/+. Among these, heterozygosity of the Hippo pathway effector yki in RpS3/+ animals resulted in the most severe morphological defects in their wings (Fig. 1D,D′, compare with 1A,A′, quantified in 1E; representative wings for the quantification are shown in Fig. S1A,B). Importantly, yki/+ or RpS3/+ flies develop normal wings (Fig. 1B-C′, quantified in 1E), indicating a specific genetic interaction between yki/+ and RpS3/+. We also found that the heterozygosity of yki causes wing morphological defects in other M/+ animals, such as RpL21/+, RpS17/+, RpS26/+ and RpS27/+ flies (Fig. S1C-F,C′-F′, quantified in S1C″-F″). Furthermore, the small-wing phenotype caused by yki knockdown was significantly enhanced by RpS3 heterozygosity (Fig. S2). These results suggest that Yki activity is required for normal wing development in M/+ flies. The wing defects in yki/+, RpS3/+ double-mutant flies were abolished when RpS3 was overexpressed in the whole wing pouch, the oval domain of the wing disc that becomes wing blade in the adult, or at the posterior compartment of the wing disc (Fig. 1F,G, quantified in 1E), suggesting that the wing defect is caused mainly by morphogenetic perturbation at the posterior compartment of yki/+, RpS3/+ wing discs.

Fig. 1.

Yki is required for normal wing development in RpS3/+ mutants. (A-D′) Adult flies (females) of wild type (wt, A), yki/+ (B) and RpS3/+ (C), and their wings (A′-D′) are shown. No morphological defects were observed except for the wings of yki/+, RpS3/+ flies (D,D′), which had smaller wings lacking the posterior wing margin. (E) Frequency of wing size for each genotype. Average wing size of the wild type (A′) was regarded as control (1.0) and then relative size to control was calculated for other genotypes. Data were analyzed using Fisher's exact test (***P<0.001; n=75, 142, 52, 208, 22, 56, respectively, for groups shown left to right). (F,G) Adult fly wings of yki/+, RpS3/+ double mutants overexpressing RpS3 in the whole pouch (F) or the posterior compartment (G).

Fig. 1.

Yki is required for normal wing development in RpS3/+ mutants. (A-D′) Adult flies (females) of wild type (wt, A), yki/+ (B) and RpS3/+ (C), and their wings (A′-D′) are shown. No morphological defects were observed except for the wings of yki/+, RpS3/+ flies (D,D′), which had smaller wings lacking the posterior wing margin. (E) Frequency of wing size for each genotype. Average wing size of the wild type (A′) was regarded as control (1.0) and then relative size to control was calculated for other genotypes. Data were analyzed using Fisher's exact test (***P<0.001; n=75, 142, 52, 208, 22, 56, respectively, for groups shown left to right). (F,G) Adult fly wings of yki/+, RpS3/+ double mutants overexpressing RpS3 in the whole pouch (F) or the posterior compartment (G).

yki/+, RpS3/+ double-mutant wing discs exhibit developmental defects caused by elevated cell death

We next examined how yki/+, RpS3/+ double mutation causes wing morphological defects. Immunostaining with anti-cleaved Dcp-1, the Drosophila effector caspase, revealed that there were only a few dying cells in wild-type and yki/+ wing discs (Fig. 2A,B, quantified in 2E) whereas there were increased levels of cell death along the dorsoventral (D/V) boundary in the RpS3/+ wing pouch (Fig. 2C, quantified in 2E) (Akai et al., 2021). Notably, massive cell death was induced along the anteroposterior (A/P) boundary in the wing discs of yki/+, RpS3/+ animals (Fig. 2D,D′, quantified in 2E). Massive cell death in yki/+, RpS3/+ wing discs along the A/P boundary was also observed by anti-cleaved caspase-3 staining (Fig. S3), which visualizes the activity of the Drosophila initiator caspase Dronc (Fan and Bergmann, 2010), suggesting that Dronc is activated at the A/P boundary of the mutant discs. Blocking cell death by overexpression of a dominant-negative form of Dronc (DroncDN) in the entire wing pouch significantly suppressed the wing defects in yki/+, RpS3/+ flies (Fig. 2F, quantified in 2G). These data suggest that Dronc-mediated ectopic cell death in the yki/+, RpS3/+ wing pouch causes developmental defects. Notably, the RpS3/+ wing pouch showed increased cell proliferation rate as visualized by anti-phospho-histone H3 staining, which indicates compensatory cell proliferation triggered by dying cells (Akai et al., 2021; Ryoo and Bergmann, 2012) (Fig. S4). However, the yki/+, RpS3/+ double-mutant wing pouch exhibited a decrease in the cell proliferation rate (Fig. S4), suggesting that the growth defects in RpS3/+, yki/+ discs are caused by both increased cell death and reduced capacity to initiate compensatory growth.

Fig. 2.

yki/+, RpS3/+ double-mutant wings cause developmental defects as a result of elevated cell death. (A-D) Wing discs of third instar larvae of the indicated genotypes stained with anti-cleaved Dcp-1 (white). Arrowheads indicate ectopic cell death (D). Scale bars: 50 µm. (D′) yki/+, RpS3/+ double mutant's wing disc expressing GFP (green) via the ptc-Gal4 driver. Ectopic cell death (magenta; indicated by arrowheads) was induced in the posterior compartment. Scale bar: 50 µm. (E) Quantification of the number of Dcp-1-positive cells per pouch. Data were analyzed using the Steel–Dwass test (***P<0.001; n=19, 20, 16, 22, respectively, for groups shown left to right). Here, and in subsequent figures, box-and-whisker plots represent the median (black line), first and third quartiles (box), and minimum and maximum within 1.5 interquartile ranges (whisker). (F) Adult wing of a yki/+, RpS3/+ double-mutant fly expressing a dominant-negative form of Dronc. (G) Frequency of wing size for each genotype. Data were analyzed using Fisher's exact test (***P<0.001; n=208, 93, respectively, for groups shown left to right).

Fig. 2.

yki/+, RpS3/+ double-mutant wings cause developmental defects as a result of elevated cell death. (A-D) Wing discs of third instar larvae of the indicated genotypes stained with anti-cleaved Dcp-1 (white). Arrowheads indicate ectopic cell death (D). Scale bars: 50 µm. (D′) yki/+, RpS3/+ double mutant's wing disc expressing GFP (green) via the ptc-Gal4 driver. Ectopic cell death (magenta; indicated by arrowheads) was induced in the posterior compartment. Scale bar: 50 µm. (E) Quantification of the number of Dcp-1-positive cells per pouch. Data were analyzed using the Steel–Dwass test (***P<0.001; n=19, 20, 16, 22, respectively, for groups shown left to right). Here, and in subsequent figures, box-and-whisker plots represent the median (black line), first and third quartiles (box), and minimum and maximum within 1.5 interquartile ranges (whisker). (F) Adult wing of a yki/+, RpS3/+ double-mutant fly expressing a dominant-negative form of Dronc. (G) Frequency of wing size for each genotype. Data were analyzed using Fisher's exact test (***P<0.001; n=208, 93, respectively, for groups shown left to right).

yki/+, RpS3/+ double-mutant wing discs exhibit cell death caused by elevated JNK signaling

We next investigated the upstream mechanism that induces ectopic cell death in yki/+, RpS3/+ wing discs. Notably, activity of JNK signaling, which can cause cell death in a variety of cellular contexts in the imaginal discs (Igaki, 2009), was strongly elevated in yki/+, RpS3/+ wing discs and moderately elevated in RpS3/+ wing discs compared with wild-type or yki/+ discs, as revealed by the JNK activity reporter TRE-dsRed (Chatterjee and Bohmann, 2012) (Fig. 3A-D″, quantified in 3E). We confirmed this by examining expression of the JNK target MMP1 (Uhlirova and Bohmann, 2006) (Fig. S5A-D, quantified in Fig. S5E and Fig. S8) and the JNK activity reporter puc-lacZ (Fig. S5F-I, quantified in S5J). The JNK activation coincided with cell death (Fig. 3D′, Fig. S3D″), suggesting that cell death is induced by JNK activation along the A/P boundary (Fig. S5D′). Indeed, blocking JNK signaling by overexpression of the JNK phosphatase Puc (Martín-Blanco et al., 1998) or heterozygosity of the bsk (JNK) (Riesgo-Escovar et al., 1996), hep (JNKK) (Glise et al., 1995) or Tak1 (JNKKK) (Mihaly et al., 2001) genes significantly reduced the extent of cell death in yki/+, RpS3/+ wing discs (Fig. 3F-I, Fig. S5O-R, quantified in Fig. S3L and Fig. S5V) and significantly suppressed wing defects in yki/+, RpS3/+ double mutants (Fig. 3M-P, quantified in 3S). Furthermore, reducing Drosophila TNF signaling (which activates the JNK pathway) by heterozygosity of eiger (egr, a TNF homolog) (Igaki et al., 2002) or knockdown of grindelwald (grnd, a TNFR homolog) (Andersen et al., 2015) also suppressed cell death in yki+, RpS3/+ wing discs (Fig. 3J,K, Fig. S5K,S-U, quantified in Fig. 3L and Fig. S5L,V) and significantly suppressed wing defects in yki/+, RpS3/+ double mutants (Fig. 3Q,R, Fig. S5M, quantified in Fig. 3S and Fig. S5N). Intriguingly, blocking Dronc activity by DroncDN overexpression also suppressed JNK activation in yki+, RpS3/+ wing discs (Fig. 3T, quantified in 3U), suggesting the presence of a signal amplification loop between JNK and Dronc (Shlevkov and Morata, 2012) (Fig. 3V).

Fig. 3.

yki/+, RpS3/+ double-mutant wing discs cause cell death via elevated JNK signaling. (A-D) Wing discs of third instar larvae of the indicated genotypes expressing TRE-dsRed (JNK signaling reporter; white). Asterisks indicate endogenous JNK activation. Arrowheads indicate JNK activation along the A/P boundary (C,D-D″). Scale bars: 50 µm. (E) Illustration of a wing disc showing the areas where mean fluorescence intensities were measured, the formula used to calculate mean intensity in the pouch relative to that in the hinge, and quantification of mean intensity in the pouch relative to that in the hinge. Data were analyzed using the Steel–Dwass test (***P<0.001; n=21, 24, 21, 20, respectively, for groups shown left to right). (F-K) Wing discs of third instar larvae of the indicated genotypes stained with anti-cleaved Dcp-1 (white). Scale bars: 50 µm. (L) Quantification of the number of Dcp-1-positive cells per pouch. Data were analyzed using the Mann–Whitney U-test (**P<0.01, ***P<0.001; n=22, 22, 19, 32, 39, 50, 21, 19, respectively, for groups shown left to right). (M-R) Adult wings of the indicated genotypes. (S) Frequency of wing size for each genotype. Data were analyzed using Fisher's exact test (**P<0.01, ***P<0.001; n=208, 97, 35, 60, 178, 58, 79, respectively, for groups shown left to right). (T) A wing disc of third instar larva of the indicated genotype stained with anti-MMP1 (white). Scale bar: 50 µm. (U) Quantification of the number of MMP1-positive cells per pouch. Data were analyzed using the Mann–Whitney U-test (**P<0.01; n=27, 22, respectively, for groups shown left to right). (V) A model for the Dronc-JNK amplification loop in yki/+, RpS3/+ double mutants.

Fig. 3.

yki/+, RpS3/+ double-mutant wing discs cause cell death via elevated JNK signaling. (A-D) Wing discs of third instar larvae of the indicated genotypes expressing TRE-dsRed (JNK signaling reporter; white). Asterisks indicate endogenous JNK activation. Arrowheads indicate JNK activation along the A/P boundary (C,D-D″). Scale bars: 50 µm. (E) Illustration of a wing disc showing the areas where mean fluorescence intensities were measured, the formula used to calculate mean intensity in the pouch relative to that in the hinge, and quantification of mean intensity in the pouch relative to that in the hinge. Data were analyzed using the Steel–Dwass test (***P<0.001; n=21, 24, 21, 20, respectively, for groups shown left to right). (F-K) Wing discs of third instar larvae of the indicated genotypes stained with anti-cleaved Dcp-1 (white). Scale bars: 50 µm. (L) Quantification of the number of Dcp-1-positive cells per pouch. Data were analyzed using the Mann–Whitney U-test (**P<0.01, ***P<0.001; n=22, 22, 19, 32, 39, 50, 21, 19, respectively, for groups shown left to right). (M-R) Adult wings of the indicated genotypes. (S) Frequency of wing size for each genotype. Data were analyzed using Fisher's exact test (**P<0.01, ***P<0.001; n=208, 97, 35, 60, 178, 58, 79, respectively, for groups shown left to right). (T) A wing disc of third instar larva of the indicated genotype stained with anti-MMP1 (white). Scale bar: 50 µm. (U) Quantification of the number of MMP1-positive cells per pouch. Data were analyzed using the Mann–Whitney U-test (**P<0.01; n=27, 22, respectively, for groups shown left to right). (V) A model for the Dronc-JNK amplification loop in yki/+, RpS3/+ double mutants.

DIAP1 expression is downregulated in yki+, RpS3/+ double mutants

We then sought to identify the Yki-target gene(s) responsible for adult wing defects in yki/+, RpS3/+ mutant animals. A strong candidate for the target is DIAP1, which inhibits apoptosis by ubiquitylating and degrading caspases, including Dronc. DIAP1 is ubiquitously expressed throughout the wing disc (Fig. 4D,D′). Intriguingly, whereas yki/+ or RpS3/+ mutants showed a similar DIAP1 expression pattern to that of wild type (Fig. 4D-F′, quantified in 4H,I), the DIAP1 levels were significantly reduced in the posterior compartment in yki/+, RpS3/+ double-mutant wing discs (Fig. 4G,G′, quantified in 4H,I). This reduction in DIAP1 expression in the posterior compartment could be the cause of the massive cell death observed at the A/P boundary in yki/+, RpS3/+ wing discs. Indeed, overexpression of DIAP1 in the entire wing pouch or the posterior compartment of the wing disc significantly blocked cell death (Fig. 4J,K, quantified in 4L), JNK activation (Fig. 4J′,K′, quantified in 4M) and wing defects (Fig. 4A,B, quantified in 4C) in yki/+, RpS3/+ mutants.

Fig. 4.

DIAP1 expression is downregulated in yki+, RpS3/+ double mutants. (A,B) Adult wings of yki/+, RpS3/+ double-mutant flies expressing DIAP1 in the whole pouch (A) or the posterior compartment (B). (C) Frequency of wing size for each genotype. Data were analyzed using Fisher's exact test (***P<0.001; n=208, 40, 66, respectively, for groups shown left to right). (D-G′) Wing discs of third instar larvae of the indicated genotypes expressing GFP (green) via the ci-Gal4 driver (D-G) and stained with anti-DIAP1 (white) (D′-G′). Scale bars: 50 µm. (H) Ratios of the fluorescence intensity in the posterior compartment relative to the anterior in D′-G′. Data were analyzed using the Steel–Dwass test (**P<0.01, ***P<0.001; n=22, 14, 20, 18, respectively, for groups shown left to right). (I) Average fluorescence intensities of anti-DIAP1-staining along the D/V boundary (indicated area in D′; cyan). Horizontal dotted lines indicate average intensities in the anterior compartment for each genotype. Vertical dotted lines indicate the A/P boundary. Note that DIAP1 expression level is downregulated in the posterior compartment in yki/+, RpS3/+ double mutants. n=22, 14, 20, 18, respectively, for groups shown left to right. (J,K) Wing discs of third instar larvae of the indicated genotypes stained with anti-cleaved Dcp-1 (white). Scale bars: 50 µm. (J′,K′) Wing discs of third instar larvae of the indicated genotypes stained with anti-MMP1 (white). Scale bars: 50 µm. (L) Quantification of the number of Dcp-1-positive cells per pouch. Data were analyzed using the Mann–Whitney U-test (***P<0.001; n=22, 28, 13, respectively, for groups shown left to right). (M) Quantification of the number of MMP1-positive cells per pouch. Data were analyzed using the Mann–Whitney U-test (***P<0.001; n=27, 19, 12, respectively, for groups shown left to right).

Fig. 4.

DIAP1 expression is downregulated in yki+, RpS3/+ double mutants. (A,B) Adult wings of yki/+, RpS3/+ double-mutant flies expressing DIAP1 in the whole pouch (A) or the posterior compartment (B). (C) Frequency of wing size for each genotype. Data were analyzed using Fisher's exact test (***P<0.001; n=208, 40, 66, respectively, for groups shown left to right). (D-G′) Wing discs of third instar larvae of the indicated genotypes expressing GFP (green) via the ci-Gal4 driver (D-G) and stained with anti-DIAP1 (white) (D′-G′). Scale bars: 50 µm. (H) Ratios of the fluorescence intensity in the posterior compartment relative to the anterior in D′-G′. Data were analyzed using the Steel–Dwass test (**P<0.01, ***P<0.001; n=22, 14, 20, 18, respectively, for groups shown left to right). (I) Average fluorescence intensities of anti-DIAP1-staining along the D/V boundary (indicated area in D′; cyan). Horizontal dotted lines indicate average intensities in the anterior compartment for each genotype. Vertical dotted lines indicate the A/P boundary. Note that DIAP1 expression level is downregulated in the posterior compartment in yki/+, RpS3/+ double mutants. n=22, 14, 20, 18, respectively, for groups shown left to right. (J,K) Wing discs of third instar larvae of the indicated genotypes stained with anti-cleaved Dcp-1 (white). Scale bars: 50 µm. (J′,K′) Wing discs of third instar larvae of the indicated genotypes stained with anti-MMP1 (white). Scale bars: 50 µm. (L) Quantification of the number of Dcp-1-positive cells per pouch. Data were analyzed using the Mann–Whitney U-test (***P<0.001; n=22, 28, 13, respectively, for groups shown left to right). (M) Quantification of the number of MMP1-positive cells per pouch. Data were analyzed using the Mann–Whitney U-test (***P<0.001; n=27, 19, 12, respectively, for groups shown left to right).

Wing defects in yki/+, RpS3/+ flies result from downregulation of Diap1

We next tested whether downregulation of Diap1 is sufficient to cause wing defects in RpS3/+ mutants. Indeed, heterozygosity for the Diap1 gene (also known as thread; th) or knocking down of Diap1 in either the entire wing pouch or the posterior region of the wing disc induced wing defects in RpS3/+ mutants without yki mutation (Fig. 5A-F, quantified in 5G). In contrast, knockdown of Diap1 in the anterior compartment of the wing discs did not cause wing defects in RpS3/+ flies (Fig. S6), indicating that Diap1 expression in the posterior region is crucial for normal wing development in RpS3/+ mutants. In all cases, JNK activation and cell death were induced in the wing pouch (Fig. 5H-M′, quantified in 5N,O). Notably, the cell proliferation rate in the RpS3/+ Diap1-RNAi wing pouch was significantly higher than that in the yki/+, RpS3/+ wing pouch (Fig. S7), suggesting that mild growth defects in RpS3/+ Diap1-RNAi wings was due to the effect of compensatory cell proliferation. Together, these data suggest that loss of one copy of yki in RpS3/+ mutants causes reduction in DIAP1 expression in the posterior compartment of the wing disc, which triggers Dronc-JNK amplification and subsequent caspase-dependent cell death.

Fig. 5.

Wing defects in yki/+, RpS3/+ flies result from downregulation of Diap1. (A,B) Adult wings of th/+ (Diap1/+) (A) and th/+, RpS3/+ (B) double-mutant flies. (C,D) Adult wings in which Diap1 has been knocked down in the whole pouch of the wing disc (C) and RpS3/+ (D) flies. (E,F) Adult wings of the indicated genotypes. (G) Frequency of wing size for each genotype. Data were analyzed using Fisher's exact test (*P<0.05, **P<0.01, ***P<0.001; n=131, 342, 90, 120, 82, 84, respectively, for groups shown left to right). (H-M) Wing discs of third instar larvae of the indicated genotypes stained with anti-cleaved Dcp-1 (white). Arrowheads indicate ectopic cell death (K,M). Scale bars: 50 µm. (H′-M′) Wing discs of third instar larvae of the indicated genotypes stained with anti-MMP1 (white). Arrowheads indicate JNK activation (K′,M′). Scale bars: 50 µm. (N) Quantification of the number of Dcp-1-positive cells per pouch. Data were analyzed using the Steel–Dwass test (*P<0.05, ***P<0.001; n=19, 13, 16, 14, 19, 17, 16, 19, 19, 13, 16, 27, respectively, for groups shown left to right). (O) Quantification of the number of MMP1-positive cells per pouch. Data were analyzed using the Steel–Dwass test (***P<0.001; n=26, 18, 19, 17, 26, 14, 19, 25, 26, 19, 19, 17, respectively, for groups shown left to right).

Fig. 5.

Wing defects in yki/+, RpS3/+ flies result from downregulation of Diap1. (A,B) Adult wings of th/+ (Diap1/+) (A) and th/+, RpS3/+ (B) double-mutant flies. (C,D) Adult wings in which Diap1 has been knocked down in the whole pouch of the wing disc (C) and RpS3/+ (D) flies. (E,F) Adult wings of the indicated genotypes. (G) Frequency of wing size for each genotype. Data were analyzed using Fisher's exact test (*P<0.05, **P<0.01, ***P<0.001; n=131, 342, 90, 120, 82, 84, respectively, for groups shown left to right). (H-M) Wing discs of third instar larvae of the indicated genotypes stained with anti-cleaved Dcp-1 (white). Arrowheads indicate ectopic cell death (K,M). Scale bars: 50 µm. (H′-M′) Wing discs of third instar larvae of the indicated genotypes stained with anti-MMP1 (white). Arrowheads indicate JNK activation (K′,M′). Scale bars: 50 µm. (N) Quantification of the number of Dcp-1-positive cells per pouch. Data were analyzed using the Steel–Dwass test (*P<0.05, ***P<0.001; n=19, 13, 16, 14, 19, 17, 16, 19, 19, 13, 16, 27, respectively, for groups shown left to right). (O) Quantification of the number of MMP1-positive cells per pouch. Data were analyzed using the Steel–Dwass test (***P<0.001; n=26, 18, 19, 17, 26, 14, 19, 25, 26, 19, 19, 17, respectively, for groups shown left to right).

Our present study shows that the Hippo pathway effector Yki plays a crucial role in ensuring robust tissue growth of Drosophila ribosomal protein mutants. In the wing imaginal epithelium of RpS3/+ mutants, endogenous Yki activity prevents Dronc-JNK-mediated cell death at the A/P boundary via the Yki target DIAP1, thereby preventing morphogenesis defects (Fig. 6). Intriguingly, in the yki/+, RpS3/+ wing pouch, JNK activation was observed not only in the posterior compartment cells but also in some cells anterior to the A/P boundary (Fig. S5). Given that the region just anterior to the A/P boundary also plays a role in regulating wing size (Willsey et al., 2016), fine-tuning of JNK activity around the A/P boundary could be crucial for robust morphogenesis. Our genetic data revealed that Egr-Grnd is essential for the induction of cell death in the yki/+, RpS3/+ wing pouch. Although the mechanism by which Egr-Grnd is activated in the mutant tissue is currently unknown, the source of Egr could be the wing disc (Igaki et al., 2009), fat body (Agrawal et al., 2016) or hemocytes (Cordero et al., 2010), and Grnd could be activated by elevated endocytosis (Igaki et al., 2009; Palmerini et al., 2021). Intriguingly, although cell death was localized to the A/P boundary in the yki/+, RpS3/+ double mutant wing pouch, adult wing defects were observed in the posterior part of the wing, suggesting some non-autonomous effect caused by local induction of massive cell death in the wing tissue.

Fig. 6.

A model for robust tissue growth via Yki-DIAP1. In RpS3/+ wing discs, JNK signaling is potentially elevated along the A/P boundary but is not amplified because the JNK-Dronc feedback loop is suppressed by the Yki-DIAP1 axis. In yki/+, RpS3/+ double mutants, JNK-Dronc amplification occurs due to DIAP1 downregulation, leading to the induction of ectopic cell death and morphogenesis defects.

Fig. 6.

A model for robust tissue growth via Yki-DIAP1. In RpS3/+ wing discs, JNK signaling is potentially elevated along the A/P boundary but is not amplified because the JNK-Dronc feedback loop is suppressed by the Yki-DIAP1 axis. In yki/+, RpS3/+ double mutants, JNK-Dronc amplification occurs due to DIAP1 downregulation, leading to the induction of ectopic cell death and morphogenesis defects.

Heterozygous mutations in ribosomal protein genes are associated with a group of diverse human diseases called ribosomopathies. Given that molecules identified in our study in Drosophila are conserved throughout evolution, Yki-mediated robust morphogenesis may be an evolutionarily conserved program for preventing developmental abnormalities in ribosomal protein mutants. Given that Yki positively regulates Myc expression (Ziosi et al., 2010), which enhances ribosomal RNA synthesis (Grewal et al., 2005), a part of the synergistic induction of cell death in RpS3/+, yki/+ double mutant discs could be triggered by a global reduction in ribosomal function. It is intriguing that morphological defects are observed only in wings of yki/+, RpS3/+ mutants. Wing morphology, the posterior part of which is more flexible in size (Imasheva et al., 1995), is crucial for insects' survival strategies, as wing shape and structure affect flight performance, foraging and courtship behavior, thus contributing to exploration and exploitation of new habitats (Beldade et al., 2011). Our findings that endogenous Yki activity in the RpS3/+ wing pouch ensures robust morphogenesis may provide a mechanistic explanation for the plastic changes in wing shape that enable adaptation to a new environment.

Fly strains

The following Drosophila strains were obtained from the Bloomington Drosophila Stock Center: RpS3Plac92 (#5627), RpS174 (#6358), TRE-DsRed (#59012), UAS-DIAP1 (#6657), bsk1 (#3088), hepr75 (#6761), UAS-Dcr.2 (#24650), UAS-Dcr.2 (#24651), th4 (#5053), pucE69 (#6762); Drosophila Genomics and Genetic Resources: RpS26KG00230 (#114620), M(3)96C2 (#101668), M(2)39F1 (#107282); the National Institute of Genetics: UAS-Diap1 RNAi (#12284R-2); and the Vienna Drosophila Resource Center: UAS-Grnd RNAi (#v43534), UAS-Grnd RNAi (#v104538), UAS-yki RNAi (#v104523). Additional strains used were: ykiB5 (a gift from D. J. Pan, The University of Texas Southwestern Medical Center, USA), egr1 (a gift from Masayuki Miura, The University of Tokyo, Japan), UAS-CD8-PARP-Venus (a gift from Yash Hiromi, National Institute of Genetics, Japan), UAS-DroncDN (a gift from Sharad Kumar, University of South Australia, Australia), UAS-Puc (a gift from Dr Enrique Martin-Blanco, IBMB - Molecular Biology Institute of Barcelona, Spain), nub-Gal4 [Bloomington Drosophila Stock Center (BDSC)], ci-Gal4 (a gift from Dr Robert Holmgren, Northwestern University, USA), en-Gal4 (a gift from Dr Erina Kuranaga, Tohoku University, Japan), ptc-Gal4 (BDSC), UAS-RpS3 (Akai et al., 2018), UAS-GFP (BDSC), tak11 (a gift from Dr Bruno Lemaitre, EPFL, Lausanne, Switzerland), w1118 (BDSC). All experimental flies carried a single copy of the transgenes. All flies were raised in vials containing standard cornmeal-glucose-yeast food and maintained at 25°C except for the experiments in Fig. S2B,C, in which flies were crossed and incubated at 18°C.

See supplementary Materials and Methods for full details of the genotypes used in each figure.

Immunohistochemistry

Wandering third instar larvae were dissected and larval tissues were fixed with 4% paraformaldehyde for 20 min at room temperature and permeabilized with 0.1% Triton X-100 in PBS for 1 h. After blocking with 5% donkey serum and 0.1% Triton X-100 solution, samples were incubated at 4°C overnight with primary antibodies, and then incubated with Alexa Fluor 488-, 546-, or 647-conjugated secondary antibodies (1:250; Invitrogen) for 2 h at room temperature. Samples were incubated and mounted with SlowFade Gold Antifade Reagent (Invitrogen) containing 4′,6-diamidino-2-phenylindole (DAPI). Primary antibodies used were: rabbit anti-cleaved caspase-3 (D175) (1:100; Cell Signaling Technology, 9661S), rabbit anti-cleaved Drosophila Dcp-1 (Asp216) (1:100; Cell Signaling Technology, 9578S), mouse anti-Wg [1:500; Developmental Studies Hybridoma Bank (DSHB)], mouse anti-MMP1 (1:100; from 1:1:1 cocktail of 3A6B4, 3B8D12 and 5H7B11; DSHB), mouse anti-DIAP1 (1:100 for 21 h; a gift from Bruce Hay, California Institute of Technology, USA), rabbit anti-phospho-histone H3 (Ser10) (1:100; Cell Signaling Technology, 9701S), mouse anti-phospho-histone H3 (Ser10) (6G3) (1:100; Cell Signaling Technology, 9706S), chicken anti-β-galactosidase (1:1000; Abcam, 9361). Images were taken with confocal microscopes (TCS-SP5, Leica Microsystems; LSM880 with Airyscan, Carl Zeiss).

Genetic screen

A series of 2nd and 3rd chromosomal deficiency lines obtained from the Bloomington Drosophila Stock Center, as well as mutant fly lines that were not uncovered by the series of deficiencies (322 lines in total) were crossed with RpS3/+ strains and phenotypes of F1 flies were examined.

Quantification and statistical analysis

For quantifying adult wing size, female adult wings were imaged with a Leica DFC 310 FX camera along with Leica FireCam software at a constant magnification and then wing blade size was manually measured using NIH ImageJ software by creating a polygon selection as indicated in magenta in Fig. S1A. Average wing size of w1118 was regarded as wild-type control (1.0) and then size relative to the control was calculated for other genotypes. For quantifying cell number, single confocal sections that possessed the largest amount of signal per sample were used in all experiments except for PH3 staining. For counting PH3-positive cells, z-stacked images were used. The number of cells positive for each antibody in the pouch region was calculated using NIH ImageJ software. Pouch region in the wing disc was determined by the expression of Wg or GFP under the nubbin-Gal4 driver. For quantifying TRE-dsRed and puc-lacZ, single confocal sections that exhibited the largest amount of signal per sample were used. For each image, mean fluorescence intensities in the pouch and the hinge regions were measured using NIH ImageJ software, as shown in Fig. 3E, and the ratio of mean fluorescence intensity in the pouch to that in the hinge was calculated. For comparing DIAP1 expression levels, single confocal sections were used and a fluorescence intensity profile was obtained by measuring the fluorescence intensity of DIAP1-expressing cells along the D/V boundary using NIH ImageJ software. To calculate the ratio of the fluorescence intensity in the posterior compartment relative to the anterior, single confocal sections were analyzed for each genotype. Mean intensity in each compartment was measured and the posterior mean intensity was divided by that of the anterior compartment. All statistical analyses were performed using R (version 3.5.2) software. Details of statistical evaluations and the numbers of samples are indicated in the figure legends. No statistical methods were used to predetermine sample size. All n numbers represent biological replicates. Each experiment was independently performed at least three times. Experiments were not randomized or blinded.

RT-qPCR

Total RNA was extracted from ten wing discs of wandering third instar larvae using the NucleoSpin RNA XS kit (Machery-Nagel). Extracted RNA was eluted in 20 µl RNase-free water and used to generate cDNA using SuperScript IV Reverse Transcriptase (Thermo Fisher Scientific). Synthesized cDNA was used for qPCR reactions using THUNDERBIRD SYBR qPCR Mix (TOYOBO) on a StepOnePlus Real-Time PCR System (Applied Biosystems). PCR conditions were 95°C for 60 s and 40 cycles of 95°C for 15 s and 60°C for 30 s. RpL32 was used as a reference gene and relative gene expression was calculated using the comparative Ct method.

Primer sequences (5′-3′) used were as follows: RpL32 (Willis et al., 2010) forward CCAAGATCGTGAAGAAGCG, reverse GTTGGGCATCAGATACTGTC; Mmp1 (Toggweiler et al., 2016) forward GAAGGCTCGGACAACGAGT, reverse GTCGTTGGACTGGTGATCG.

We thank K. Baba, M. Koijima, M. Tanaka, M. Matsuoka, K. Gomi and Y. Kobe for technical assistance; B. Hay for sharing anti-DIAP1 antibody; D. Pan, J. Jiang, M. Miura, Y. Hiromi, S. Kumar, the Bloomington Stock Center (BDSC, Indiana, USA), the Vienna Drosophila RNAi Center (VDRC, Vienna, Austria), the National Institute of Genetics Stock Center (NIG, Shizuoka, Japan) and the Drosophila Genetic Resource Center (DGRC, Kyoto Stock Center, Japan) for fly stocks. We also thank members of the Igaki laboratory for discussions.

Author contributions

Conceptualization: Y.W., S.O., T.I.; Methodology: Y.W., S.O., T.I.; Validation: Y.W., T.I.; Formal analysis: Y.W., S.O., T.I.; Investigation: Y.W., S.O.; Resources: Y.W., S.O., T.I.; Writing - original draft: Y.W.; Writing - review & editing: Y.W., S.O., T.I.; Supervision: S.O., T.I.; Project administration: S.O., T.I.; Funding acquisition: S.O., T.I.

Funding

This work was supported by KAKENHI grants from the Ministry of Education, Culture, Sports, Science and Technology/Japan Society for the Promotion of Science (16H02505, 19K22424 and 20H05945 to S.O. and T.I.), the Naito Foundation (S.O. and T.I.), the Takeda Science Foundation (S.O. and T.I.), Inamori Foundation (S.O.), Toray Science Foundation (S.O.), Senri Life Science Foundation (S.O.) and the Mitsubishi Foundation (S.O.).

The peer review history is available online at https://journals.biologists.com/dev/article-lookup/doi/10.1242/dev.198705

Agrawal
,
N.
,
Delanoue
,
R.
,
Mauri
,
A.
,
Basco
,
D.
,
Pasco
,
M.
,
Thorens
,
B.
and
Léopold
,
P.
(
2016
).
The Drosophila TNF Eiger is an Adipokine that acts on insulin-producing cells to mediate nutrient response
.
Cell Metab.
23
,
675
-
684
.
Akai
,
N.
,
Igaki
,
T.
and
Ohsawa
,
S.
(
2018
).
Wingless signaling regulates winner/loser status in Minute cell competition
.
Genes Cells
23
,
234
-
240
.
Akai
,
N.
,
Ohsawa
,
S.
,
Sando
,
Y.
and
Igaki
,
T.
(
2021
).
Epithelial cell-turnover ensures robust coordination of tissue growth in Drosophila ribosomal protein mutants
.
PLoS Genet.
17
,
e1009300
.
Andersen
,
D. S.
,
Colombani
,
J.
,
Palmerini
,
V.
,
Chakrabandhu
,
K.
,
Boone
,
E.
,
Röthlisberger
,
M.
,
Toggweiler
,
J.
,
Basler
,
K.
,
Mapelli
,
M.
,
Hueber
,
A. O.
et al.
(
2015
).
The Drosophila TNF receptor Grindelwald couples loss of cell polarity and neoplastic growth
.
Nature
522
,
482
-
486
.
Aspesi
,
A.
and
Ellis
,
S. R.
(
2019
).
Rare ribosomopathies: insights into mechanisms of cancer
.
Nat. Rev. Cancer
19
,
228
-
238
.
Beldade
,
P.
,
Mateus
,
A. R. A.
and
Keller
,
R. A.
(
2011
).
Evolution and molecular mechanisms of adaptive developmental plasticity
.
Mol. Ecol.
20
,
1347
-
1363
.
Bergmann
,
A.
(
2010
).
The role of ubiquitylation for the control of cell death in Drosophila
.
Cell Death Differ.
17
,
61
-
67
.
Brehme
,
K. S.
(
1939
).
A study of the effect on development of “minute” mutations in Drosophila melanogaster
.
Genetics
24
,
131
-
13161
.
Chatterjee
,
N.
and
Bohmann
,
D.
(
2012
).
A versatile ϕC31 based reporter system for measuring AP-1 and NRF2 signaling in Drosophila and in tissue culture
.
PLoS ONE
7
,
e34063
.
Cmejla
,
R.
,
Cmejlova
,
J.
,
Handrkova
,
H.
,
Petrak
,
J.
and
Pospisilova
,
D.
(
2007
).
Ribosomal protein S17 gene (RPS17) is mutated in Diamond-Blackfan anemia
.
Hum. Mutat.
28
,
1178
-
1182
.
Cordero
,
J. B.
,
Macagno
,
J. P.
,
Stefanatos
,
R. K.
,
Strathdee
,
K. E.
,
Cagan
,
R. L.
and
Vidal
,
M.
(
2010
).
Oncogenic ras diverts a host TNF tumor suppressor activity into tumor promoter
.
Dev. Cell
18
,
999
-
1011
.
Ditzel
,
M.
,
Broemer
,
M.
,
Tenev
,
T.
,
Bolduc
,
C.
,
Lee
,
T. V.
,
Rigbolt
,
K. T. G.
,
Elliott
,
R.
,
Zvelebil
,
M.
,
Blagoev
,
B.
,
Bergmann
,
A.
et al.
(
2008
).
Inactivation of effector caspases through nondegradative polyubiquitylation
.
Mol. Cell
32
,
540
-
553
.
Doherty
,
L.
,
Sheen
,
M. R.
,
Vlachos
,
A.
,
Choesmel
,
V.
,
O'Donohue
,
M. F.
,
Clinton
,
C.
,
Schneider
,
H. E.
,
Sieff
,
C. A.
,
Newburger
,
P. E.
,
Ball
,
S. E.
et al.
(
2010
).
Ribosomal protein genes RPS10 and RPS26 are commonly mutated in diamond-Blackfan anemia
.
Am. J. Hum. Genet.
86
,
222
-
228
.
Fan
,
Y.
and
Bergmann
,
A.
(
2010
).
The cleaved-Caspase-3 antibody is a marker of Caspase-9-like DRONC activity in Drosophila
.
Cell Death Differ.
17
,
534
-
539
.
Farley-Barnes
,
K. I.
,
Ogawa
,
L. M.
and
Baserga
,
S. J.
(
2019
).
Ribosomopathies: Old concepts, new controversies
.
Trends Genet.
35
,
754
-
767
.
Glise
,
B.
,
Bourbon
,
H.
and
Noselli
,
S.
(
1995
).
. hemipterous encodes a novel drosophila MAP kinase kinase, required for epithelial cell sheet movement
.
Cell
83
,
451
-
461
.
Grewal
,
S. S.
,
Li
,
L.
,
Orian
,
A.
,
Eisenman
,
R. N.
and
Edgar
,
B. A.
(
2005
).
Myc-dependent regulation of ribosomal RNA synthesis during Drosophila development
.
Nat. Cell Biol.
7
,
295
-
302
.
Igaki
,
T.
(
2009
).
Correcting developmental errors by apoptosis: lessons from Drosophila JNK signaling
.
Apoptosis
14
,
1021
-
1028
.
Igaki
,
T.
,
Kanda
,
H.
,
Yamamoto-Goto
,
Y.
,
Kanuka
,
H.
,
Kuranaga
,
E.
,
Aigaki
,
T.
and
Miura
,
M.
(
2002
).
Eiger, a TNF superfamily ligand that triggers the Drosophila JNK pathway
.
EMBO J.
21
,
3009
-
3018
.
Igaki
,
T.
,
Pastor-Pareja
,
J. C.
,
Aonuma
,
H.
,
Miura
,
M.
and
Xu
,
T.
(
2009
).
Intrinsic tumor suppression and epithelial maintenance by endocytic activation of Eiger/TNF signaling in Drosophila
.
Dev. Cell
16
,
458
-
465
.
Imasheva
,
A. G.
,
Bubli
,
O. A.
,
Lazebny
,
O. E.
and
Zhivotovsky
,
L. A.
(
1995
).
Geographic differentiation in wing shape in Drosophila melanogaster
.
Genetica
96
,
303
-
306
.
Kongsuwan
,
K.
,
Yu
,
Q.
,
Vincent
,
A.
,
Frisardi
,
M. C.
,
Rosbash
,
M.
,
Lengyel
,
J. A.
and
Merriam
,
J.
(
1985
).
A Drosophila Minute gene encodes a ribosomal protein
.
Nature
317
,
555
-
558
.
Lee
,
T. V.
,
Fan
,
Y.
,
Wang
,
S.
,
Srivastava
,
M.
,
Broemer
,
M.
,
Meier
,
P.
and
Bergmann
,
A.
(
2011
).
Drosophila IAP1-mediated ubiquitylation controls activation of the initiator caspase DRONC independent of protein degradation
.
PLoS Genet.
7
,
e1002261
.
Martín-Blanco
,
E.
,
Gampel
,
A.
,
Ring
,
J.
,
Virdee
,
K.
,
Kirov
,
N.
,
Tolkovsky
,
A. M.
and
Martinez-Arias
,
A.
(
1998
).
. puckered encodes a phosphatase that mediates a feedback loop regulating JNK activity during dorsal closure in Drosophila
.
Genes Dev.
12
,
557
-
670
.
Marygold
,
S. J.
,
Roote
,
J.
,
Reuter
,
G.
,
Lambertsson
,
A.
,
Ashburner
,
M.
,
Millburn
,
G. H.
,
Harrison
,
P. M.
,
Yu
,
Z.
,
Kenmochi
,
N.
,
Kaufman
,
T. C.
et al.
(
2007
).
The ribosomal protein genes and Minute loci of Drosophila melanogaster
.
Genome Biol.
8
,
R216
.
Mihaly
,
J.
,
Kockel
,
L.
,
Gaengel
,
K.
,
Weber
,
U.
,
Bohmann
,
D.
and
Mlodzik
,
M.
(
2001
).
The role of the Drosophila TAK homologue dTAK during development
.
Mech. Dev.
102
,
67
-
79
.
Misra
,
J. R.
and
Irvine
,
K. D.
(
2018
).
The hippo signaling network and its biological functions
.
Annu. Rev. Genet.
52
,
65
-
87
.
Palmerini
,
V.
,
Monzani
,
S.
,
Laurichesse
,
Q.
,
Loudhaief
,
R.
,
Mari
,
S.
,
Cecatiello
,
V.
,
Olieric
,
V.
,
Pasqualato
,
S.
,
Colombani
,
J.
,
Andersen
,
D. S.
et al.
(
2021
).
Drosophila TNFRs Grindelwald and Wengen bind Eiger with different affinities and promote distinct cellular functions
.
Nat. Commun.
12
,
2070
.
Riesgo-Escovar
,
J. R.
,
Jenni
,
M.
,
Fritz
,
A.
and
Hafen
,
E.
(
1996
).
The Drosophila jun-N-terminal kinase is required for cell morphogenesis but not for DJun-dependent cell fate specification in the eye
.
Genes Dev.
10
,
2759
-
2768
.
Ryoo
,
H. D.
and
Bergmann
,
A.
(
2012
).
The role of apoptosis-induced proliferation for regeneration and cancer
.
Cold Spring Harbor Perspect. Biol.
4
,
a008797
.
Shlevkov
,
E.
and
Morata
,
G.
(
2012
).
A dp53/JNK-dependent feedback amplification loop is essential for the apoptotic response to stress in Drosophila
.
Cell Death Differ.
19
,
451
-
460
.
Toggweiler
,
J.
,
Willecke
,
M.
and
Basler
,
K.
(
2016
).
The transcription factor Ets21C drives tumor growth by cooperating with AP-1
.
Sci. Rep.
6
,
1
-
10
.
Uhlirova
,
M.
and
Bohmann
,
D.
(
2006
).
JNK- and Fos-regulated Mmp1 expression cooperates with Ras to induce invasive tumors in Drosophila
.
EMBO J.
25
,
5294
-
5304
.
Wang
,
R.
,
Yoshida
,
K.
,
Toki
,
T.
,
Sawada
,
T.
,
Uechi
,
T.
,
Okuno
,
Y.
,
Sato-Otsubo
,
A.
,
Kudo
,
K.
,
Kamimaki
,
I.
,
Kanezaki
,
R.
et al.
(
2015
).
Loss of function mutations in RPL27 and RPS27 identified by whole-exome sequencing in Diamond-Blackfan anaemia
.
Br. J. Haematol.
168
,
854
-
864
.
Willis
,
D. K.
,
Wang
,
J.
,
Lindholm
,
J. R.
,
Orth
,
A.
and
Goodman
,
W. G.
(
2010
).
Microarray analysis of juvenile hormone response in Drosophila melanogaster S2 cells
.
J. Insect Sci.
10
,
66
.
Willsey
,
H. R.
,
Zheng
,
X.
,
Carlos Pastor-Pareja
,
J.
,
Willsey
,
A. J.
,
Beachy
,
P. A.
and
Xu
,
T.
(
2016
).
Localized JNK signaling regulates organ size during development
.
eLife
5
,
e11491
.
Ziosi
,
M.
,
Baena-López
,
L. A.
,
Grifoni
,
D.
,
Froldi
,
F.
,
Pession
,
A.
,
Garoia
,
F.
,
Trotta
,
V.
,
Bellosta
,
P.
,
Cavicchi
,
S.
and
Pession
,
A.
(
2010
).
dMyc Functions downstream of yorkie to promote the supercompetitive behavior of hippo pathway mutant cells
.
PLoS Genet.
6
,
e1001140
.

Competing interests

The authors declare no competing or financial interests.

Supplementary information