Structural distinctions in BMPs underlie divergent signaling in spinal neurons
© Perron and Dodd; licensee BioMed Central Ltd. 2012
Received: 17 February 2012
Accepted: 4 April 2012
Published: 4 May 2012
In dorsal spinal neurons and monocytes, bone morphogenetic protein (BMP)7 activates distinct transduction pathways, one leading to inductive specification and the other to axon orientation and chemotaxis. BMP7-evoked induction, also stimulated by the closely related BMP6, acts through a Smad cascade, leading to nuclear signaling, and is not BMPR subunit selective. Orientation is evoked by BMP7, but not by BMP6, through PI3K-dependent cytoskeletal activation mediated by the type II BMPRs, ActRIIA and BMPRII and is independent of the Smad cascade. The responses can be stimulated concurrently and suggest that BMP7, but not BMP6, can selectively activate BMPR subunits that engage the divergent paths. Although structural and biochemical analyses of selected BMP/BMPR interfaces have identified key regions of interaction, how these translate into function by related BMPs is poorly understood. To determine the mechanisms underlying the distinct activities of BMP7 and the disparate properties of BMP7 and BMP6 in spinal cord development, we have performed a family-wide structure/function analysis of BMPs and used the information to predict and test sites within BMPs that may control agonist properties, in particular the ability of a BMP to orient axons, through interactions with BMPRs.
We demonstrate that whereas all BMPs can induce dorsal neurons, there is selectivity in the ability also to orient axons or evoke growth cone collapse. The degree to which a BMP orients is not predictable by overall protein similarity with other BMPs but comparison of sequences of potent and weakly orienting BMPs with that of the non-orienting BMP6 revealed three candidate positions within the BMPs at which the amino acid residues may confer or obstruct orienting ability. Residue swapping analysis has identified one residue, Gln 48 in BMP6, that blocks axon orienting ability. Replacing Gln 48 with any of the amino acids present at the equivalent residue position in the orienting subset of BMPs confers orienting activity on BMP6. Conversely, swapping Gln 48 into BMP7 reduces orienting ability. The inductive capacity of the BMPs was unchanged by these residue swaps.
The results suggest that the presence of the Gln 48 residue in BMP6 is structurally inhibitory for BMP/BMPR interactions that result in the activation of intracellular signaling, leading to axon orientation. Moreover, since residue 48 in BMP7 and the corresponding residue in BMP2 are important for type II BMPR binding, our results provide a basis for a mechanistic understanding of the diverse activities of BMPs in spinal cord development.
Bone morphogenetic proteins (BMPs) represent a class of TFGβ factors with diverse functions in mammals. BMPs evoke transcriptional events leading to cellular differentiation and survival [1, 2] but also direct axon guidance and cellular orienting activities through cytoskeletal signaling [3–5]. The large number of BMPs, their differing affinities for an array of heteromeric BMP receptors (BMPRs), and a range of extracellular and intracellular modulators of BMP/BMPR interactions [6, 7] suggest that differential expression of these BMP signaling components could produce the many actions of BMPs. However, in neurons and monocytes, a single BMP can simultaneously evoke both transcriptional and cytoskeletal responses that involve different receptor subunits and divergent intracellular signaling programs [3, 8, 9], suggesting additional mechanisms that supersede the simple distribution of transduction components. Moreover, the ability to activate divergent pathways is not necessarily shared by closely related BMPs, highlighting the importance of individual agonist properties [8, 9]. We sought here to understand the properties that underlie the selective ability of a BMP to exert divergent orienting activity.
In the mammalian central nervous system (CNS), BMP6, BMP7 and the more distantly related GDF7, are expressed with overlapping distribution in the roof plate, at the dorsal midline of the developing spinal cord [4, 10]. Roof plate BMP activity mediates the induction of dorsal spinal interneurons (dI neurons) [10, 11] and the subsequent guidance of the axons of nascent dI neurons but examination of the individual BMPs uncovered different roles for these family members during spinal cord development [4, 12]. Comparison of the inductive and orienting activities of BMP6, BMP7 and GDF7 in spinal explants revealed that while all three induce ectopic dI neurons, only BMP7 can also orient extending dI axons [4, 10, 12]. Moreover, whereas BMP6 stimulates induction but has no orienting activity in dissociated dI neurons or monocytes, BMP7 activates inductive signaling and also evokes growth cone collapse in dI neurons and chemotaxis in monocytes [4, 9, 12]. Nonetheless, the different activities of the closely related BMP6 and BMP7 in dorsal spinal development have remained a puzzle.
The results summarized above, combined with the finding that orienting responses to BMP7 are initiated at much lower concentrations than BMP-evoked inductive signaling in the same cells [3, 8, 9], led to the idea that, whereas both BMP7 and BMP6 engage receptor complexes that activate intracellular inductive machinery, BMP7 alone recruits a distinct receptor complex that directs signaling towards the cytoskeleton. In support of this, we and others have provided evidence that the type II BMPR subunits, ActRIIA and BMPRII, and the type I BMPR subunit, BMPRIB, are required selectively by BMP7 to activate cellular orienting responses of neurons and monocytes but are not individually essential for BMP inductive activity [8, 13]. In recent work, both BMP7 and BMP6 have been shown to activate Smad signaling and dI neuron induction through the activity of type I BMPR kinases, whereas BMP7-evoked axonal orientation is independent of type I BMPR kinase activity . Furthermore, BMP7, but not BMP6, activates a PI3K-dependent pathway required for axon orientation but PI3K-dependent signaling is not required for BMP7-stimulated induction . Together, these findings suggest a model whereby the ability of BMP7 to stimulate divergent intracellular pathways within a cell results from the differential recruitment and/or activation of BMPRs, but a mechanistic explanation for such selective subunit recruitment is still lacking.
BMPs bind as dimers to BMPR complexes comprising one pair each of type I BMPR and type II BMPR subunits . Classically, phosphorylation of the type I receptors by the type II receptor pair leads to stimulation of the intracellular cascade of activated Smads [1, 14] resulting in transcriptional signaling. In addition, however, type II BMPRs appear to interact directly with signaling intermediates that regulate cytoskeletal events [15, 16]. Variation in receptor composition and the mode of receptor subunit recruitment by different BMPs are thought to contribute to choice of distinct intracellular pathways [17, 18] and this notion, combined with differences in binding potencies of BMPs to BMPRs [19–21], suggests that individual BMPs may achieve distinct cellular outcomes through differential usage of subunits in BMPR complexes.
An unresolved issue, however, is the basis for selective agonist interactions with these BMPR subunits. Structural modeling and binding studies of BMP interactions with the extracellular domains (ECD) of BMPRs have identified sites of interaction between select BMPs and type I and type II BMPRs [22–24]. Indeed, the predicted BMP7/ActRIIA and BMP2/BMPRIA interfaces are known and found in spatially distinct regions [23, 25, 26]. It seems likely that understanding how these regions of the BMPs contribute to their agonist properties will elucidate the mechanisms underlying selective BMP function in developing spinal cord. We have, therefore, performed a structure/function analysis across the BMP family, examining inductive, growth cone collapsing and axon orienting ability on dI neurons. We have used this information to identify features of the primary structure of BMPs that may control orienting ability and have begun to test these inferences using a residue swapping approach. Our analysis reveals one residue within the putative type II BMPR binding domain of BMP7 and BMP6 which is critical for orienting ability, while not impinging on inductive capacity. These results suggest a model for the mechanism underlying the ability of BMPs to recruit or activate selectively certain BMPR subunits leading to axon orientation signaling.
Characterization of orienting activity across the BMP family
The ability of the BMPs and TGFβs to induce differentiation of ectopic dI neurons was assessed by measuring ectopic Lhx2/9 expression (Figure 3B) in the same set of [d] explants used to record axon orientation. In control [d] explants, cultured adjacent to pellets of COS-1 cells expressing pMT23, endogenous expression of Lhx2/9 was restricted to dorsal regions of the explants (Figure 3B and see ). In [d] explants co-cultured with BMP7-expressing COS-1 cells ectopic Lhx2/9 expression was observed (Figure 3B). The D-V extent of Lhx2/9 expression was expanded in response to BMP7, reflecting a 2.7-fold increase in dI1 neurons (Figure 3C (red bars)). All 10 BMPs showed similarly robust dI1-inducing ability (Figure 3B, C) stimulating induction of Lhx2/9+ cells in a range of 153 to 215% over control. None of the other TGFβs tested induced ectopic Lhx2/9 expression (Figure 3C). These results indicate that although all BMPs have dI1 neuron inducing capacity, only a subset exhibits axon orienting activity.
Discrete structural distinctions correlate with the ability of BMPs to orient axons
BMPs fall into closely related subgroups assigned according to overall mature protein similarity (see Figure 3D). Comparison of subgroup relationships with respect to orienting activity of BMPs (compare Figure 3C and D) revealed that axon orienting BMPs are not those most highly related by overall amino acid similarities. BMP7, BMP6 and BMP5 represent the most closely interrelated subset of BMPs (Figure 3D) yet display orienting activity that can be categorized as high, insignificant and intermediate, respectively (Figure 3C (green bars)). BMP9 is relatively distantly related to BMP7 yet has the most comparable orienting ability. The capacity to induce ectopic Lhx2/9 expression was similar across the family of BMPs but did not extend to other members of the TGFβ family (Figure 3C (red bars)). Thus the ability of BMPs to induce neural character appears to be a property common to all BMPs, but the ability of BMPs to orient is restricted to a subset of active BMPs that does not mirror groupings according to overall structural similarity.
To probe further the identity of residues that might confer orienting activity, we next compared the sequences of BMP6 and BMP7 with the sequences of the other BMPs with robust orienting activity. Of the eight amino acid residues that are expressed in BMP6 but not BMP7, four are also expressed in other robustly orienting BMPs (green boxes in Figure 4B) and one (residue 36) contains a conservative substitution between BMP6 and BMPs2/9. Thus these five residues seem unlikely to represent positions critical for orienting activity. The three remaining divergent residues, at positions 48, 65 and 98 in BMP6, are not present in the corresponding positions in any of the BMPs with orienting ability, including BMP5, designating these residues as candidate determinants of orienting ability. We therefore next examined the influence of residues 48, 65 and 98 in BMP6 and BMP7 to confer or depress orienting activity.
Residue swapping selectively alters BMP orienting ability
We first examined the abilities of BMP6/BMP7 chimeras to evoke dI neuron growth cone collapse, comparing them with the activities of BMP6 WT, BMP7 WT and rBMP7. Dissociated dI neurons in sister cultures were incubated with either COS-1 cell CM or rBMP7 and changes in growth cone area were assessed as in Figure 1B, C. BMP7 WT CM caused reduction of growth cone area that was not significantly different from that evoked by rBMP7 (41 ± 4.0% BMP7 WT; 37 ± 0.2% rBMP7, compared to control CM from COS-1 cell cultures expressing vector alone; Figure 5B). BMP6 WT CM did not cause significant growth cone collapse (12 ± 2.3% reduction in growth cone area; Figure 5B). Similarly, BMP6 N65Y and BMP6 Y98T CM were ineffective in evoking growth cone collapse, showing no improvement over BMP6 WT CM (13 ± 4.9% and 20 ± 3.1% reduction in growth cone area, respectively; Figure 5B). In contrast, BMP6 Q48R CM caused a decrease in growth cone area (50 ± 4.4% reduction in growth cone area; Figure 5B) that was similar to that stimulated by BMP7 WT CM and rBMP7. The dramatic acquisition of activity by BMP6 Q48R but not by BMP6 N65Y or BMP6 Y98T suggested that residue 48 is important for the ability of a BMP to evoke growth cone collapse.
Glutamine at position 48 depresses orienting activity
The ability of BMP6 Q48R to orient axons and evoke growth cone collapse suggests either that there is a requirement for Arg 48 (which is large and positively charged) or that Gln 48 (which is large and uncharged) is inhibitory to the selective BMP to BMPR interactions needed for orientation activity. Inspection of the residues found in other orienting BMPs, BMP2, BMP4 and BMP9, at the position corresponding to Arg 48 in BMP7 suggests that the properties of Arg 48 may not be explicitly required to control orientation activity: BMP2 and BMP4 show serine (Ser) (which is small and uncharged) instead of arginine and BMP9 contains glutamic acid (Glu) (which is of medium size and negatively charged) at this position (Figure 4B). Nonetheless, Arg, Ser and Glu may each confer orienting activity. To test this, we first asked whether swapping Ser or Glu into position 48 in BMP6 would confer orienting activity. Both BMP6 Q48S and BMP6 Q48E showed strong orienting activity (33 ± 1.4° and 27 ± 1.4°, respectively; Figure 6A, B and not shown), appearing as effective as BMP6 Q48R and BMP7 WT. As with BMP6 Q48R, both BMP6 Q48S and BMP6 Q48E showed similar inductive activity to that of BMP6 WT (Figure 6A, C).
To determine whether the acquisition of orienting activity by BMP6 Q48R, Q48S and Q48E chimeras reflects a requirement specifically for any one of the three amino acids found in other orienting BMPs, we replaced Gln 48 in BMP6 with a generic amino acid, alanine (A). BMP6 Q48A evoked robust growth cone collapse (47 ± 5.0% reduction in growth cone area; Figure 5B) and oriented dI axons within [d] explants (40 ± 2.4°; Figure 6B), showing similar responses to those evoked by BMP6 Q48R. Together, these results suggest that the beneficial effect of residue swapping into position 48 may result from removal of Gln 48 rather than a selective positive effect of Arg, Ser or Glu. The inability of BMP6 to orient might be determined by Gln 48 preventing interaction between BMP6 and a BMPR at a critical interface. To test the possibility that the Gln 48 residue negatively regulates activity in BMP6 we therefore generated a reciprocal BMP7 construct in which Arg 48 in BMP7 was replaced with Gln (BMP7 R48Q). Strikingly, BMP7 R48Q had substantially reduced growth cone collapse activity (27 ± 5.7% reduction in growth cone area; Figure 5B) and dI axon repulsion (19 ± 2.4°; Figure 6A, B) by comparison with BMP7 WT activity, whereas inductive activity in the same explants was unchanged (Figure 6A, C). These results suggest that Gln 48 determines the inability of BMP6 to orient axons.
Here we have addressed an unexplained aspect of agonist specificity in BMP signaling, a feature that may control the different functions of highly related BMPs during early spinal cord development. We have explored the mechanism by which two highly related BMPs have dramatically different abilities to activate signaling that regulates cytoskeletal dynamics, leading to axon orientation, although sharing the ability to evoke inductive signaling in the same cell. That these distinct responses can be activated by a single BMP and appear to depend on different receptor subunit activation led us to examine the BMP family for agonist properties that might influence selective BMP/BMPR interactions. We show that the activities of the whole family recapitulate the disparate functions of the roof plate BMPs. Whereas all BMPs share the ability to induce dI neuron differentiation, only a subset of the family have axon orienting activity and this subset is not predicted by overall sequence similarity. Exploiting the close similarities of the BMP family, however, we identified amino acid residues as candidates for conferring orienting ability on BMPs. Generation of chimeric BMPs by single amino acid substitutions has revealed a critical position at which the residue confers or reduces orienting activity but does not influence inductive activity. Although this critical residue is unlikely to be the sole determinant of BMP orienting activity, it lies within the predicted interface between BMP7 and ActRIIA, a receptor required for orienting activity. These results provide mechanistic insight into the ability of BMPs to recruit distinct receptor complexes and elicit different functional outcomes.
Comparative analysis reveals residue 48 as critical for orienting activity
Identification of BMPs as “orienting” or “non-orienting” was fundamental to the structure/function analysis presented here and two lines of evidence provided confidence that all BMPs were tested well within the functional concentration range for orientation. First, in Western blots of COS-1 cell lysates and CM similar levels of efficient expression were observed for all native and chimeric constructs, indicating that residue swapping did not interfere with BMP processing or secretion. Second, in COS-1 cell CM, BMPs were expressed in 25-fold excess of the concentration required for BMP7 to evoke orientation . We and others have shown that while neural induction and activation of associated downstream signaling components requires high concentrations of BMPs, the orienting activities of BMP7, including growth cone collapse and monocyte chemotaxis, occur at considerably lower concentrations [3, 4, 9]. The finding that all 10 native BMPs and each chimeric BMP induced the ectopic expression of Lhx2/9 in [d] explants, and to a similar extent, suggests that explants were exposed to comparable concentrations of BMPs. Thus the observed absence of axon orienting activity in BMP6 WT and BMP6 N65Y and Y98T chimeras, as well as the reduction in axon orienting activity in the BMP7 R48Q chimera, reflects an inability to activate the relevant pathways through specific agonist properties and leads to the conclusion that residue 48 in BMP6 and BMP7 influences receptor binding in a critical manner.
Residue 48 is unlikely to be the sole determinant of orienting ability
Single amino acid substitutions in BMPs have previously been shown to cause dramatic differences in BMP activities and in BMPR binding [28–30]. The identification of the residue at position 48 as critical for the orienting activity of BMPs adds to this cannon. However, in experiments with the reciprocal chimera, BMP7 R48Q, orienting activity is reduced but not fully inhibited, indicating that residue 48 in BMP6/7 is not the only determinant of orienting activity, and that there may be subtle mitigating effects from other residues present in BMP7 but not BMP6. The existence of additional determinants is also suggested by our finding that orienting activity across the BMP family is not simply categorized in binary fashion. BMP5, most closely related in sequence to BMP7 and BMP6, and containing an arginine residue (Arg 47 ) at the position corresponding to Arg 48 in BMP7, typifies a group of BMPs with intermediate orienting activity in [d] explants. In this functional study, we chose first to examine residues in BMP6 and BMP7 that were the most different from each other (non-conservative amino acid substitutions). However, considering both conservative and non-conservative amino acid differences observed in BMP6 and BMP7 sequences, several are matched between the sequences of BMP5 and BMP6. These residues are located mainly in a region thought to be important for type I BMPR binding [23, 26] and therefore may not be involved directly in specific type II BMPR interactions. Nonetheless, these and other differences present in the BMP sequences may influence the conformation of the BMP ligand and its interaction with the BMPR complex, such that graduated functional outcomes are generated. The intermediate orienting activity of BMP5 may therefore be a useful tool, in conjunction with BMP6 and BMP7, with which to identify amino acids that can confer BMP6-like or BMP7-like activity upon BMP5 and to provide a more comprehensive picture of the determinants for orienting activity mediated by BMPs.
How does residue 48 affect BMP agonist properties?
Crystal structure studies of BMP7 bound to ActRIIA show that two of the BMP7 residues that differ in BMP6, Arg 48 and Tyr 65 , lie within the predicted BMP7/ActRIIA interface [22, 23, 26]. These two residues thus represent candidate regulators of selective interactions with ActRIIA, one of the BMPR subunits that we have shown to be required for orienting activity of BMP7 . Replacing residue 65 in BMP6 with Tyr, however, appeared not to confer orienting activity on BMP6, but coordinate multiple residue swaps might be necessary to reveal a more subtle role for residue position 65. Notably, none of the BMP6 or BMP7 chimeras altered BMP-evoked inductive signaling, revealing a singular role for residue 48 and representing a point of agonist/receptor interaction required selectively for orientation. However, our results throw into question the idea that specific amino acids are required to be present at position 48 in BMP7, BMP2/4 and BMP9 for orienting activity. Rather, Arg Ser and Glu may be permissive for this activity, whereas the properties of Gln 48 prevent BMP6 from activating the signaling apparatus that transduces orienting activity. Indeed, the potencies of BMP6 Q48R, Q48S, Q48E and Q48A in axon orientation were as high as that of BMP9, the most potent of the BMPs, whereas in BMP7 swapping of Arg 48 to Gln 48 reduced orienting activity dramatically.
Several studies have explored the importance of the mode of binding of BMPs to receptor subunits. The cellular response to BMPs has been shown to depend on the mode of receptor oligomerization [17, 18, 34]. BMP binding to preformed receptor complexes drives Smad-dependent, transcriptional pathways, whereas BMP-induced receptor subunit assembly leads to non-transcriptional responses, such as cytoskeletal rearrangements [17, 18]. Although the presence of the type I BMPR, BMPRIB, is necessary for the orientation response , type I BMPR kinase activity, which is required for BMP-evoked Smad phosphorylation and inductive signaling, appears not to be important for orientation . In contrast, BMP7-evoked monocyte chemotaxis requires the selective engagement of type II BMPRs, ActRIIA and BMPRII . Together, these data suggest a model (Figure 7) in which BMP7 and other orienting BMPs are able selectively to recruit a receptor complex responsible for BMP orienting activity, presumably comprising ActRIIA, BMPRII and BMPRIB (Figure 7E), whereas BMP6 is unable to stimulate orienting activity through this receptor complex (Figure 7E), perhaps prevented by structural differences in which the Gln 48 residue plays a key role in type II BMPR binding specificity. In contrast, BMP-evoked inductive activity (Figure 7F), which appears to be common to all BMPs but requires higher agonist concentrations and may reflect less stringent requirements for BMPR binding, perhaps involving preformed receptors.
BMPs share the ability to induce the differentiation of dorsal spinal neurons, but only a subset exhibit orientation activity toward the axons of these neurons. The closely related BMP family members, BMP6 and BMP7 with 95% similarity, most dramatically illustrate this difference. BMP7 does and BMP6 does not display orienting activity. We show here that a single amino acid residue at position 48 in BMP7 is a major determinant of BMP orienting activity. Residue swapping of Gln 48 in BMP6 and Arg 48 present in BMP7 confers robust axon orienting ability upon BMP6 and reduces activity in BMP7. Moreover, replacing the Gln 48 residue in BMP6 with the corresponding residue in other orienting BMPs permitted BMP6 to orient axons and collapse growth cones of dI neurons. In contrast, none of these manipulations altered BMP dI inductive activity. Our results suggest that the presence of the Gln 48 residue in BMP6 is structurally inhibitory for transduction of the signals necessary for axon orientation by BMPs and provide a basis for our mechanistic understanding of the diverse activities of BMPs in spinal cord development.
Materials and methods
Antibodies and reagents
Recombinant BMPs were purchased from R&D Systems, Minneapolis, MN, USA, and stock solutions were prepared in 4 mM HCl/0.1% BSA. Antibodies: mouse α-TAG-1 (4D7; ), rabbit α-Lhx2/9 (L1; ), mouse α-ERM (13 H9; ), rabbit α-HA, mouse α-HA (Abcam, Cambridge, MA, USA), mouse α-myc (9E10; ), rabbit α-BMP9 (see below). HRP- and fluorophore-conjugated secondary antibodies were obtained from Jackson ImmunoResearch Laboratories, West Grove, PA, USA. Cell culture reagents: Ham’s F12 medium, Opti-MEM medium, Penicillin/Streptomycin/Glutamine (P/S/G), Penicillin/Streptomycin (P/S) (Invitrogen, Carlsbad, CA, USA), FBS (Gemini BioProducts, West Sacramento, CA, USA) and 45% glucose (Sigma-Aldrich, St Louis, MO, USA). Pharmacological reagent: LY294002 (LY; Cell Signaling Technology, Danvers, MA, USA): stock solution was prepared in DMSO and subsequently diluted in serum-free Opti-MEM/P/S/G.
The α-BMP9 polyclonal antibody was generated by immunizing rabbits with a peptide (MGVPTLKYHYEG) corresponding to C-terminal amino acids 133 through 144 in the mature coding sequence of mouse BMP9 (Covance Research Products, Inc., Denver, PA, USA). The antiserum was affinity purified using the Montage Antibody Purification Kit (Millipore Corporation, Bedford, MA, USA).
Generation of BMP/GDF/TGFβ expression constructs
Mouse cDNA was prepared from whole E11.5 embryos with Superscript II reverse transcriptase (Invitrogen). The mature regions of all BMP/TGFβ superfamily members used in this study were generated by PCR using gene-specific primers containing either myc (EQKLISEEDL), flag (DYKDDDDK) or HA (YPYDVPDYA) epitope tag insertions and cloned into pMT23 as previously described [10, 38].
Single amino acid mutations were performed using the QuickChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA, USA). Primers were designed and used to produce nucleotide changes in the native HA-tagged mouse BMP6.pMT23 expression construct to yield the Q48R, Q48A, Q48S, Q48E, N65Y and Y98T chimeras and in the native HA-tagged mouse BMP7.pMT23 construct to yield the R48Q chimera. The sequences were verified by analysis (Genewiz, South Plainfield, NJ, USA).
Generation and characterization of conditioned medium
COS-1 cells plated in 35 mm dishes were transfected with control vector (pMT23), epitope-tagged native or chimeric BMP expression constructs by transfection with either Lipofectamine Reagent or Lipofectamine LTX Reagent (Invitrogen). Following an overnight incubation, the medium was replaced with 2 mL serum-free Opti-MEM/P/S/G. Conditioned medium (CM) was collected after 48 hours and used in growth cone collapse assays. For analysis of secreted BMPs, CM was concentrated 20-fold with Centricon YM-3 Amicon concentrators (Millipore, Bedford, MA, USA). Concentrated CM or whole cell lysates, prepared from transfected COS-1 cells using 1x Lysis Buffer (Cell Signaling Technology,) supplemented with 1 mM PMSF, were separated by SDS-PAGE (EZ-Run Gel Solution, Fisher Scientific, Pittsburgh, PA, USA) and transferred to nitrocellulose (Whatman, Clifton, NJ, USA). Nitrocellulose membranes were blocked in 5% non-fat milk/0.1% Tween 20/TBS (Blocking Buffer) and probed overnight with epitope tag or BMP-specific antibodies diluted in Blocking Buffer. Membranes were washed in TBST (0.1% Tween 20/TBS) and probed (1 hour) with HRP-conjugated secondary antibodies in Blocking Buffer. After washing in TBST, blots were developed using the Supersignal West Pico chemiluminescent substrate detection kit (Pierce, Rockford, IL, USA) and exposed to Kodak BioMax Light Film Kodak, Rochester, NY, USA).
Growth cone collapse assay
Dissociated dI neuron cultures were prepared as previously described . The cultures were serum-starved by incubation in unsupplemented F12 medium for 2 hours at 37°C and stimulated with rBMPs at 50 ng/ml or with COS-1 cell CM (500μL) for 30 minutes. The cultures were fixed in pre-warmed 4% paraformaldehyde/0.5% gluteraldehyde/0.1 M phosphate buffer for 5 minutes, washed once in PBS, blocked in 1% heat-inactivated goat serum/0.1% Triton X-100/PBS and labeled with a mouse α-ERM IgM and a Cy3 goat α-mouse IgM secondary antibody. The growth cone area of neurons with axons greater the 10 μm was measured across two or three coverslips per condition for each experiment using ImageJ 1.37v software (National Institutes of Health (NIH)). Growth cone collapsing activity is presented as raw mean area or as the percentage decrease of growth cone area relative to control cultures.
[i] explant assays
Intermediate spinal cord ([i]) explants were dissected from stage 10 chick embryos, cultured in collagen and immunolabeled as previously described [39, 40]. BMPs (50 ng/ml) or 4 mM HCl/0.1% BSA (control) were incubated with the explants for 48 hours.
[d] explant assays
E11 rat [d] explants were dissected, cultured and labeled as previously described . COS-1 cells were transfected with epitope-tagged native or chimeric BMP expression constructs using Lipofectamine Reagent or Lipofectamine LTX Reagent (Invitrogen), aggregated  and appended to [d] explants as described . Explants were immunolabeled with antibodies against TAG-1, Lhx2/9 and the epitope tag of the COS-1 cell-expressed BMP. Lhx2/9 induction and the angle of reorientation of TAG-1+ dI axons were measured in each explant in parallel. Quantitation of Lhx2/9 induction, using ImageJ (NIH), was performed by measuring the percentage change in integrated density (mean pixel intensity x area) of the BMP-induced region of Lhx2/9+ cells present in the explant relative to control (pMT23) explants. The angle of reorientation was measured as shown previously .
Images of dI neuron dissociated cultures were taken with a Zeiss AxioCam HR digital camera (Carl Zeiss, Thornwood, NY, USA) mounted on a Zeiss Axiovert 200 M fluorescence microscope. In addition, images of [i] and [d] explants were taken using a Zeiss LSM 5 confocal microscope and are presented here as confocal Z-stacks.
Alignments and dendrogram
Amino acid alignments of the mature regions of BMPs were made using Vector NTI software (Invitrogen). Sequences are aligned beginning two amino acids upstream from the first conserved cysteine of the mature region [10, 11]). The BMP/TGFβ phylogenetic tree dendrogram was generated using ArboDraw v1.3 software (http://dunbrack.fcccc.edu/ArboDraw). All sequences are from mouse, except for chick dorsalin-1.
Bone morphogenetic protein
Bone morphogenetic protein receptor
Bovine serum albumin
COS-1 cell conditioned medium
- [d] explants:
Explant of E11 rat dorsal spinal cord
Dorsal spinal interneuron
Dorsal to ventral
Embryonic day where E0.5 = 6 am on the day of plug
Ezrin-radixin-moeisin FBS, fetal bovine serum
Human influenza hemagglutinin
- [i] explant:
Explant of intermediate region of spinal cord of Hamburger Hamilton Stage 10 chick embryo
Transient axonal glycoprotein
Transforming growth factor beta
We are grateful to Beth Shafer for collaboration in the early stages of the project, to Brie Wamsley and Georgie Nicholl for superb technical assistance, to Susan Morton for assistance in designing the α-BMP9 antibody and to Thomas Jessell for reagents. We wish to thank members of the Dodd and Jessell labs for many helpful discussions and particularly Thomas Jessell and Ming Zhou for helpful comments on the manuscript. The work was funded, in its early stages, by the Christopher and Dana Reeve Foundation (JD).
- Massague J: How cells read TGF-beta signals. Nat Rev Mol Cell Biol. 2000, 1: 169-178.View ArticlePubMedGoogle Scholar
- Liu A, Niswander LA: Bone morphogenetic protein signalling and vertebrate nervous system development. Nat Rev Neurosci. 2005, 6: 945-954. 10.1038/nrn1805.View ArticlePubMedGoogle Scholar
- Cunningham NS, Paralkar V, Reddi AH: Osteogenin and recombinant bone morphogenetic protein 2B are chemotactic for human monocytes and stimulate transforming growth factor beta 1 mRNA expression. Proc Natl Acad Sci U S A. 1992, 89: 11740-11744. 10.1073/pnas.89.24.11740.PubMed CentralView ArticlePubMedGoogle Scholar
- Augsburger A, Schuchardt A, Hoskins S, Dodd J, Butler S: BMPs as mediators of roof plate repulsion of commissural neurons. Neuron. 1999, 24: 127-141. 10.1016/S0896-6273(00)80827-2.View ArticlePubMedGoogle Scholar
- Gamell C, Osses N, Bartrons R, Ruckle T, Camps M, Rosa JL, Ventura F: BMP2 induction of actin cytoskeleton reorganization and cell migration requires PI3-kinase and Cdc42 activity. J Cell Sci. 2008, 121: 3960-3970. 10.1242/jcs.031286.View ArticlePubMedGoogle Scholar
- Miyazono K, Kamiya Y, Morikawa M: Bone morphogenetic protein receptors and signal transduction. J Biochem. 2010, 147: 35-51. 10.1093/jb/mvp148.View ArticlePubMedGoogle Scholar
- Bragdon B, Moseychuk O, Saldanha S, King D, Julian J, Nohe A: Bone morphogenetic proteins: a critical review. Cell Signal. 2011, 23: 609-620. 10.1016/j.cellsig.2010.10.003.View ArticlePubMedGoogle Scholar
- Perron JC, Dodd J: ActRIIA and BMPRII Type II BMP receptor subunits selectively required for Smad4-independent BMP7-evoked chemotaxis. PLoS One. 2009, 4: e8198-10.1371/journal.pone.0008198.PubMed CentralView ArticlePubMedGoogle Scholar
- Perron JC, Dodd J: Inductive specification and axonal orientation of spinal neurons mediated by divergent bone morphogenetic protein signaling pathways. Neural Dev. 2011, 6: 36-10.1186/1749-8104-6-36.PubMed CentralView ArticlePubMedGoogle Scholar
- Lee KJ, Mendelsohn M, Jessell TM: Neuronal patterning by BMPs: a requirement for GDF7 in the generation of a discrete class of commissural interneurons in the mouse spinal cord. Genes Dev. 1998, 12: 3394-3407. 10.1101/gad.12.21.3394.PubMed CentralView ArticlePubMedGoogle Scholar
- Liem KF, Tremml G, Jessell TM: A role for the roof plate and its resident TGFbeta-related proteins in neuronal patterning in the dorsal spinal cord. Cell. 1997, 91: 127-138. 10.1016/S0092-8674(01)80015-5.View ArticlePubMedGoogle Scholar
- Butler SJ, Dodd J: A role for BMP heterodimers in roof plate-mediated repulsion of commissural axons. Neuron. 2003, 38: 389-401. 10.1016/S0896-6273(03)00254-X.View ArticlePubMedGoogle Scholar
- Yamauchi K, Phan KD, Butler SJ: BMP type I receptor complexes have distinct activities mediating cell fate and axon guidance decisions. Development. 2008, 135: 1119-1128. 10.1242/dev.012989.View ArticlePubMedGoogle Scholar
- Derynck R, Feng XH: TGF-beta receptor signaling. Biochim Biophys Acta. 1997, 1333: F105-F150.PubMedGoogle Scholar
- Derynck R, Zhang YE: Smad-dependent and Smad-independent pathways in TGF-beta family signalling. Nature. 2003, 425: 577-584. 10.1038/nature02006.View ArticlePubMedGoogle Scholar
- Foletta VC, Lim MA, Soosairajah J, Kelly AP, Stanley EG, Shannon M, He W, Das S, Massague J, Bernard O, Soosairaiah J: Direct signaling by the BMP type II receptor via the cytoskeletal regulator LIMK1. J Cell Biol. 2003, 162: 1089-1098. 10.1083/jcb.200212060.PubMed CentralView ArticlePubMedGoogle Scholar
- Nohe A, Hassel S, Ehrlich M, Neubauer F, Sebald W, Henis YI, Knaus P: The mode of bone morphogenetic protein (BMP) receptor oligomerization determines different BMP-2 signaling pathways. J Biol Chem. 2002, 277: 5330-5338. 10.1074/jbc.M102750200.View ArticlePubMedGoogle Scholar
- Sieber C, Kopf J, Hiepen C, Knaus P: Recent advances in BMP receptor signaling. Cytokine Growth Factor Rev. 2009, 20: 343-355. 10.1016/j.cytogfr.2009.10.007.View ArticlePubMedGoogle Scholar
- Kirsch T, Nickel J, Sebald W: BMP-2 antagonists emerge from alterations in the low-affinity binding epitope for receptor BMPR-II. EMBO J. 2000, 19: 3314-3324. 10.1093/emboj/19.13.3314.PubMed CentralView ArticlePubMedGoogle Scholar
- Knaus P, Sebald W: Cooperativity of binding epitopes and receptor chains in the BMP/TGFbeta superfamily. Biol Chem. 2001, 382: 1189-1195.View ArticlePubMedGoogle Scholar
- Heinecke K, Seher A, Schmitz W, Mueller TD, Sebald W, Nickel J: Receptor oligomerization and beyond: a case study in bone morphogenetic proteins. BMC Biol. 2009, 7: 59-10.1186/1741-7007-7-59.PubMed CentralView ArticlePubMedGoogle Scholar
- Shah PK, Buslje CM, Sowdhamini R: Structural determinants of binding and specificity in transforming growth factor-receptor interactions. Proteins. 2001, 45: 408-420. 10.1002/prot.10010.View ArticlePubMedGoogle Scholar
- Greenwald J, Groppe J, Gray P, Wiater E, Kwiatkowski W, Vale W, Choe S: The BMP7/ActRII extracellular domain complex provides new insights into the cooperative nature of receptor assembly. Mol Cell. 2003, 11: 605-617. 10.1016/S1097-2765(03)00094-7.View ArticlePubMedGoogle Scholar
- Yin H, Yeh LC, Hinck AP, Lee JC: Characterization of ligand-binding properties of the human BMP type II receptor extracellular domain. J Mol Biol. 2008, 378: 191-203. 10.1016/j.jmb.2008.02.031.View ArticlePubMedGoogle Scholar
- Keller S, Nickel J, Zhang JL, Sebald W, Mueller TD: Molecular recognition of BMP-2 and BMP receptor IA. Nat Struct Mol Biol. 2004, 11: 481-488. 10.1038/nsmb756.View ArticlePubMedGoogle Scholar
- Allendorph GP, Vale WW, Choe S: Structure of the ternary signaling complex of a TGF-beta superfamily member. Proc Natl Acad Sci U S A. 2006, 103: 7643-7648. 10.1073/pnas.0602558103.PubMed CentralView ArticlePubMedGoogle Scholar
- Wilson SI, Shafer B, Lee KJ, Dodd J: A molecular program for contralateral trajectory: Rig-1 control by LIM homeodomain transcription factors. Neuron. 2008, 59: 413-424. 10.1016/j.neuron.2008.07.020.View ArticlePubMedGoogle Scholar
- Nickel J, Kotzsch A, Sebald W, Mueller TD: A single residue of GDF-5 defines binding specificity to BMP receptor IB. J Mol Biol. 2005, 349: 933-947. 10.1016/j.jmb.2005.04.015.View ArticlePubMedGoogle Scholar
- Allendorph GP, Isaacs MJ, Kawakami Y, Izpisua Belmonte JC, Choe S: BMP-3 and BMP-6 structures illuminate the nature of binding specificity with receptors. Biochemistry. 2007, 46: 12238-12247. 10.1021/bi700907k.View ArticlePubMedGoogle Scholar
- Song K, Krause C, Shi S, Patterson M, Suto R, Grgurevic L, Vukicevic S, van Dinther M, Falb D, Ten Dijke P, Alaoui-Ismaili MH: Identification of a key residue mediating bone morphogenetic protein (BMP)-6 resistance to noggin inhibition allows for engineered BMPs with superior agonist activity. J Biol Chem. 2010, 285: 12169-12180. 10.1074/jbc.M109.087197.PubMed CentralView ArticlePubMedGoogle Scholar
- Sebald W, Nickel J, Zhang JL, Mueller TD: Molecular recognition in bone morphogenetic protein (BMP)/receptor interaction. Biol Chem. 2004, 385: 697-710.View ArticlePubMedGoogle Scholar
- Griffith DL, Keck PC, Sampath TK, Rueger DC, Carlson WD: Three-dimensional structure of recombinant human osteogenic protein 1: structural paradigm for the transforming growth factor beta superfamily. Proc Natl Acad Sci U S A. 1996, 93: 878-883. 10.1073/pnas.93.2.878.PubMed CentralView ArticlePubMedGoogle Scholar
- Saremba S, Nickel J, Seher A, Kotzsch A, Sebald W, Mueller TD: Type I receptor binding of bone morphogenetic protein 6 is dependent on N-glycosylation of the ligand. FEBS J. 2008, 275: 172-183. 10.1111/j.1742-4658.2007.06187.x.View ArticlePubMedGoogle Scholar
- Isaacs MJ, Kawakami Y, Allendorph GP, Yoon BH, Izpisua Belmonte JC, Choe S: Bone morphogenetic protein-2 and −6 heterodimer illustrates the nature of ligand-receptor assembly. Mol Endocrinol. 2010, 24: 1469-1477. 10.1210/me.2009-0496.PubMed CentralView ArticlePubMedGoogle Scholar
- Dodd J, Morton SB, Karagogeos D, Yamamoto M, Jessell TM: Spatial regulation of axonal glycoprotein expression on subsets of embryonic spinal neurons. Neuron. 1988, 1: 105-116. 10.1016/0896-6273(88)90194-8.View ArticlePubMedGoogle Scholar
- Goslin K, Birgbauer E, Banker G, Solomon F: The role of cytoskeleton in organizing growth cones: a microfilament-associated growth cone component depends upon microtubules for its localization. J Cell Biol. 1989, 109: 1621-1631. 10.1083/jcb.109.4.1621.View ArticlePubMedGoogle Scholar
- Evan GI, Lewis GK, Ramsay G, Bishop JM: Isolation of monoclonal antibodies specific for human c-myc proto-oncogene product. Mol Cell Biol. 1985, 5: 3610-3616.PubMed CentralView ArticlePubMedGoogle Scholar
- Basler K, Edlund T, Jessell TM, Yamada T: Control of cell pattern in the neural tube: regulation of cell differentiation by dorsalin-1, a novel TGF beta family member. Cell. 1993, 73: 687-702. 10.1016/0092-8674(93)90249-P.View ArticlePubMedGoogle Scholar
- Yamada T, Pfaff SL, Edlund T, Jessell TM: Control of cell pattern in the neural tube: motor neuron induction by diffusible factors from notochord and floor plate. Cell. 1993, 73: 673-686. 10.1016/0092-8674(93)90248-O.View ArticlePubMedGoogle Scholar
- Liem KF, Tremml G, Roelink H, Jessell TM: Dorsal differentiation of neural plate cells induced by BMP-mediated signals from epidermal ectoderm. Cell. 1995, 82: 969-979. 10.1016/0092-8674(95)90276-7.View ArticlePubMedGoogle Scholar
- Shah SB, Skromne I, Hume CR, Kessler DS, Lee KJ, Stern CD, Dodd J: Misexpression of chick Vg1 in the marginal zone induces primitive streak formation. Development. 1997, 124: 5127-5138.PubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.