The BMP-binding protein Crossveinless 2 is a short-range, concentration-dependent, biphasic modulator of BMP signaling in Drosophila

Abstract

In Drosophila, the secreted BMP-binding protein Short gastrulation (Sog) inhibits signaling by sequestering BMPs from receptors, but enhances signaling by transporting BMPs through tissues. We show that Crossveinless 2 (Cv-2) is also a secreted BMP-binding protein that enhances or inhibits BMP signaling. Unlike Sog, however, Cv-2 does not promote signaling by transporting BMPs. Rather, Cv-2 binds cell surfaces and heparan sulfate proteoglygans and acts over a short range. Cv-2 binds the type I BMP receptor Thickveins (Tkv), and we demonstrate how the exchange of BMPs between Cv-2 and receptor can produce the observed biphasic response to Cv-2 concentration, where low levels promote and high levels inhibit signaling. Importantly, we show also how the concentration or type of BMP present can determine whether Cv-2 promotes or inhibits signaling. We also find that Cv-2 expression is controlled by BMP signaling, and these combined properties enable Cv-2 to exquisitely tune BMP signaling.

Figure 1. The nature and range of Cv-2 function

(A) Model of BMP signaling in the developing PCV. (B) Wild type adult wing. (C-C′) DSRF and pMad levels in the wild type pupal PCV. (D) Diagram of the cv-2^F1-42 allele created by targeted mutagenesis. (E) cv-2 mRNA in the pupal crossvein region. (F) cv-2^F1-42 adult wing. (G-G′) IDSRF and pMad in pupal cv-2^F1-42 PCV. (H-L) Effect of homozygous cv-2^F1-42 clones on pMad (H-K′) or DSRF (L, L′) in pupal wings. Panels show both surfaces of the PCV region with corresponding clones (green or magenta outlines) or, in J-K′, the region of clone overlap (red outlines). Wild type sides of the boundaries are marked + and mutant -. Arrows mark regions of clones with normal (green) or disrupted (red) development of the PCV. (M) Adult A9-gal4 wings. A9-gal4 is expressed throughout the pupal wing, at levels lower than that of en-gal4 (not shown). (N-P) Overexpression of a single copy (N-O) or two copies (P) of UAS-cv-2 with A9-gal4 (N-O). (Q) Overexpression of low levels of cv-2 with en-gal4 and EP-driven cv-2 (EP(2)1103). (R) overexpression of high levels of cv-2 with en-gal4 and UAS-cv-2. The posterior-specific en-gal4 driver is expressed throughout most of the ACV region, ending just posterior to L3 (Ralston and Blair, 2005).

Figure 2. Biochemical characterization of Drosophila Cv-2

(A) Diagram of dual-tagged Cv-2 proteins (6xMyc N-terminal and V5-6xHis C-terminal) showing CR and vWFD domains, and cleavage and disulfide bonding between N-terminal and C-terminal fragments. All the Cv-2 variants used below were similarly tagged with C-terminal V5-6xHis and, in some cases, N-terminal 6xMyc. B, C, E-J, L), Western analyses of tagged Cv-2. The immunoblots in (B), (C), (G), and (I) show simultaneous anti-Myc (red) and anti-V5 (green) staining; in the rest of the panels only anti-V5 staining is shown. (B and C) Dual-tagged Cv-2 produced in S2 cells (B) or in embryos after overexpression of UAS-cv-2 with da-gal4 (C). (D) Conservation of the cleavage site in Cv-2 proteins: (d-Drosophila, a-Anopheles, m-mouse, h-human, z-zebrafish, c-chicken). (E) Cleavage and mobility shifts for Cv-2 variants from S2 cell supernatant under reducing {R} and non-reducing {N} conditions, or from S2 cell pellets run under reducing conditions. The slightly faster migration of full-length Cv-2 under non-reducing conditions is likely due to conformational changes. (F) Cleavage and mobility shifts for dual-tagged wild type and uncleavable (G³⁸⁷DP-AAA) cv-2 variants overexpressed in embryos with da-gal4, run under reducing {R} and non-reducing {N} conditions. (G) IP of dual-tagged Myc-Cv-2-V5/6His with either anti-Myc (center western) or Ni (right western) co-IPs the N-terminal or C-terminal fragment of Cv-2, respectively. Control lanes show absence of precipitation of single-tagged Cv-2-V5/6His by anti-Myc or of Myc-Cv-2 by Ni beads. Full length Myc-tagged Cv-2 runs at 120kD, Cv-2-V5/6His at 110kD. (H) Cv-2 cleavage requires C³⁸³ and C⁵²⁰, but not C⁴⁰⁵, C⁵¹³, C⁵¹⁷ or C⁵⁶⁰. Predicted Cys pairings in the first half of the vWFD domain are indicated. (I) Cleaved and uncleaved forms of the dual-tagged Myc-Cv-2-V5 co-precipitate with Dpp-Flag or Gbb-Flag, but not with the anti-Flag-beads alone (-). i= input Cv-2. (J) Levels of cleaved and uncleaved Cv-2 that co-IP with Dpp-Flag, compared with input levels (highest levels shown in blue). (K) Relative amounts of cleaved and uncleaved Cv-2 that co-IP with Dpp-Flag, expressed as the percentage of the input levels. (L) Full length Cv-2, but not Cv-2-N or Cv-2-C, co-IP with Flag-tagged Dpp.

Figure 3. Cv-2 interacts with the cell surface, Dally, and the BMP type I receptor Tkv

(A and B) Binding of Cv-2 and Cv-2-C, but not Cv-N, to naïve S2 cells. i = input lanes. Binding is not influenced by the addition of Dpp (+ vs. -) (A), and occurs at both room temperature (RT) and 4°C (B). (C-E) Increased binding of V5-tagged Cv-2 and Cv-2-C, but not Cv-2-N (anti-V5 in green) from S2 cell supernatant to S2 cells overexpressing Myc-tagged Dally (anti-Myc in red). DAPI (blue) stains the nuclei of transfected and untransfected cells. (F) Cv-2 (green) co-IPs with Myc-tagged Dally (red). Black arrowheads = IgG bands. (G) Reduced extracellular accumulation of A9-gal4-driven Myc-Cv-2 (anti-Myc, red) in botv clones, marked by absence of a GFP (green) marker, in wing imaginal discs. (H) Comparison of Cv-2 binding to identical numbers naïve (N), tkv dsRNA (ds) and tkv-transfected (T) S2 cells. Anti-tubulin is shown as a loading control. On average a 50% decrease was seen in 5 dsRNA repetitions. (I) Cv-2 co-IPs with Flag-tagged Tkv. While there is some background IP of Myc-Cv-2 (green) in naïve cells (N), the level of IPi is increased 4-7 fold in cells (5 repetitions) expressing Tkv1 (T), and is not affected by the addition of Dpp or Gbb. (J) Drosophila Cv-2 co-IPs with an His-tagged chimera containing the extracellular domain of human BMPR-IB, with or without recombinant Dpp (green). (K) Dpp-Flag simultaneously IPs Myc-Cv-2 (red) and a His-tagged Fc-chimera containing the extracellular portion of BMPR-IB (green). (i- input proteins; MW-molecular weight marker).

Figure 4. Modeling the biphasic activity of Cv-2

(A) Model for cell autonomous action of Cv-2. (B) Typical results for three versions of the model shown in (A), with signaling possible via BCR only (i), equal signaling via BCR and BR (ii), or signaling via BR only (iii). (C) Model results showing a biphasic response to Cv-2 levels. See Table S1 for parameter values. (D) Model results showing how similar levels of Cv-2 overexpression can still promote signaling but suppress the effects of threefold increases in BMP. See Table S1 for parameter values. (E) Overexpression of Cv-2 suppresses the effects of Dpp and Gbb overexpression on adult wings.

Figure 5. Biphasic activity of Cv-2 is ligand dependent

(A and B) Conditions that lead to biphasic activity of Cv-2. 10,000 results for model (iii) with randomly varying parameters are shown. The X, Y, and Z axes correspond to the dimensionless affinity constants C, R, and BCR. Red dots represent biphasic solutions; green dots represent antagonistic solutions. The non-dimensional thresholds were computed by adjusting by the mean R_tot or B used in the numerical screen, which gave values of 10, 10, and 0.0316 for K^-1_C, K^-1_R, and K^-1_BCR respectively. Thresholds are shown by planes that dissect the data for K_C, K_R, and K^-1_BCR. Regions are denoted by three letters that correspond to (H)igh or (L)ow) Cv-2 affinity, receptor affinity, and BCR affinity. (A) Top view shows solutions for K_BCR > 0.0316 (dimensionless) and four regions HHL, HLL, LLL, and HLL. (B) Bottom view with four regions: HHH, HLH, LLH, and HLH. (C) Histogram shows number of biphasic, antagonistic, and total solutions and the percent of biphasic solutions in each region. (D) Binding parameters for BMP-2, BMP-4, and BMP-7 obtained from (a) Sebald et al.,(2004), (b) Hatta et al.(2000), and (c) Rentzsch et al. (2006). (E and F) Typical response curves show how the level of BR changes for increasing Cv-2 with different BMP concentrations for BMP-2 (E) and BMP-7 (F). See Table S1 for parameter values. (G and H) The effect of cv-2 transfection or cv-2 RNAi on Dpp-mediated signaling (G) or Gbb mediated signaling (H) in S2 cells. Signaling is measured by the relative levels of Flag-tagged Mad (green) and pMad (red), and is quantified in the histograms.

Figure 6. Comparison of the effects of Cv-2 variants on adult wings

(A-A′) Mild overexpression of uncleavable Cv-2 with A9-gal4. (B) Strong overexpression of uncleavable Cv-2 with en-gal4. C-E) Rescue of PCV loss in cv-2^KO1 adults by A9-gal4-driven expression of wild-type cv-2 (C) and uncleavable cv-2 (D), but not by cv-2-N (E). (F-H) Rescue of the PCV loss normally caused by either cv-2¹ (F) or overexpression of Sog (G,H) by en-gal4-driven expression of Cv-2-N.

Figure 7. Positive feedback of BMP signaling on cv-2 mRNA expression

(A-C) Refinement of cv-2 mRNA expression and anti-pMad staining in pupal wings. AP = after pupariation. (D) Expression of cv-2 in a stage 5 embryo. (E and F) Loss of cv-2 expression from regions of cv⁷⁰ (E) and gbb¹/gbb⁴ (F) pupal wings. (G) cv-2 expression after overexpression of moderate levels Gbb with A9-gal4. (H and I) cv-2 expression (H) and anti-pMad staining (I) after overexpression of high levels of Gbb in the posterior of the wing with en-gal4 (J) If Cv-2 acts as a strict antagonist there is a single intersection between the binding and the positive feedback equilibria. See Table S1 for parameter values. (K) If Cv-2 is biphasic there are multiple intersections between the binding and positive feedback and equilibria, leading to bistability. Inset shows the bistable behavior as a function of the level of BMP. Points 1 and 2 are the stable steady-states whereas point 3 is unstable. Additional analysis of the full 4D system shows the dynamic approach to the stable steady-state (Umulis et al., 2006).