Introduction: Perhaps the authors can rearrange some paragraphs to simplify the background description for the non-specialist reader. It could start with the description of Dishevelled proteins (line 56) and their modular composition (please cite some references describing the DIX and the DEP domain structures). Then, it can describe the PDZ structures, their binding properties and why they became targets for drug design to treat several diseases. The authors could also mention a couple of recent reviews describing the advances in the design of peptides and small molecule modulators of PDZ domains and the logic behind this new work.

We change the introduction as requested and include more literature regarding the design of PDZ domain inhibitors, including peptidomimetics, in particular by including a very nice review from the Strømgaard group. We discuss the scope of our work more explicitly in light of the existing data.

Results: The definition of the PDZ binding site is shown as Figure S8. The authors could consider moving this figure to the main figures (New Figure 1 panel A) and label the red and blue spheres with Dishevelled residues. Having this representation next to the complexes will facilitate the description of the binding cavity in the structures (New Figure 1 remaining panels).

Regarding the result section, we will change Figure 1 as requested by transferring the back bone cartoon shown in Figure S8 into Figure 1A, citing it in the virtual screening and NMR section.

Data presentation: The authors should include an overlay of the corresponding data points (as dot plots) in Figure 2 (Bar charts), Figure S4 and Figure S5, and please, add the n number and define the error bars (e.g. SD, SEM) in the figure legends.

We improve the data representation by including of course the definition of the error bars in the figure legends of the various plots and the n numbers. Since the other two referees were ok with the presentation of the data, suggesting even that the paper is actually quite well written, we do not plan so far to change the bar charts in Fig. 2. We think they represent the data quite clearly.

Figure S1: Please use monospaced fonts to ensure that the sequences are aligned.

Figure S1 will of course be changed to monospaced fonts (courier, as it was before), and checked before submission whether the format is kept.

Equation1 Please correct this equation.

Of course, we will correct equation 1.

Most importantly, Ji Zheng requested an NMR binding assay. We have done the requested NMR binding assay with Dvl-3, which indicated binding in the three-digit micromolar range. Employing an eight-fold excess of ligand (as commonly in this study), we observed only very small chemical shift changes, with a maximum change of still below 0.1 ppm for only one signal. Table S1 lists similar chemical shift changes (but above 0.1 ppm as an average over three signals) for ligands that bind in the three-digit micromolar range (eg. compound 2 and 6). Those three shifting signals are indicated in Fig. S2, and comparison of S2 and S8 displays a considerable difference. In S2, the highest ligand concentration was 8-fold excess. Our best binding compounds show shift changes larger 0.3 ppm as an average over three signals. The compound was checked by mass spectrometry. We will include the NMR binding assay employing 8-fold excess of the compound into Figure S8 and reword the discussion related to this compound, see reply to Minye Zhang: ‘Based on our NMR and ITC studies, the binding affinity of CBC-322338/3289-8625 to DVL-3 seems to be less than 50 micromolar, comparing CSPs from the NMR assay with those of our other compounds listed in Table S1 and considering the weak heat development in our ITC assay, which was larger than the originally reported value (10.6 +/- 1.7) (Grandy 2009) that was OBTAINED WITH A DIFFERENT METHOD.’

We would like to mention that the ITC and NMR assays above were performed by different persons in different labs here on the campus in Berlin Buch, and the compound was obtained from the EU-OPENSCREEN (an EU facility) and FMP compound library that is handled under professional conditions. The compound was obtained from the facility already dissolved in DMSO and this solution was quality-controlled by LC-MS. Independently, Hori et al. found a similar binding characteristic, also in a direct binding assay.

With respect to the inclusion of the ITC and NMR results into Table 1 we made sure that the values for Dvl-1 and Dvl-3 are determined with the same method. We would like to mention that we do not have both values in all of the cases, especially as many values are associated with compounds synthesized at intermediate stages of the project, and one value was sufficient for finding the right path. It was important to us to provide values determined by ITC for the important compounds 18-21 since it may be considered the superior method in such investigations. However, we have reported chemical shift changes (DCSP) for most of the compounds in the supplementary information (Table S1 on page 14).

This referee, Mingjie Zhang, indicated that the results associated with compound 3289-8625 should be discussed differently. We will do so, and please have also a look at the reply to Jie Zheng, the other referee, were we discuss the NMR shift assay now performed for this revision, in more detail. In reply to this request and the other, we plan to include the sentence ‘Based on NMR and ITC studies, the binding affinity of CBC-322338/3289-8625 to DVL-3 seems to be less than 50 micromolar, comparing CSPs from the NMR assay with those of our other compounds listed in Table S1 and considering the weak heat development in our ITC assay, which was larger than the originally reported value (10.6 +/- 1.7) (Grandy 2009) that was OBTAINED WITH A DIFFERENT METHOD.’ The whole subject is discussed more extensively in the reply to Jie Zheng.

Of course, we will replace R2 by R1 in lines 185 and 187.

Introduction: Perhaps the authors can rearrange some paragraphs to simplify the background description for the non-specialist reader. It could start with the description of Dishevelled proteins (line 56) and their modular composition (please cite some references describing the DIX and the DEP domain structures). Then, it can describe the PDZ structures, their binding properties and why they became targets for drug design to treat several diseases. The authors could also mention a couple of recent reviews describing the advances in the design of peptides and small molecule modulators of PDZ domains and the logic behind this new work.

We change the introduction as requested and include more literature regarding the design of PDZ domain inhibitors, including peptidomimetics, in particular by including a very nice review from the Strømgaard group. We discuss the scope of our work more explicitly in light of the existing data.

Results: The definition of the PDZ binding site is shown as Figure S8. The authors could consider moving this figure to the main figures (New Figure 1 panel A) and label the red and blue spheres with Dishevelled residues. Having this representation next to the complexes will facilitate the description of the binding cavity in the structures (New Figure 1 remaining panels).

Regarding the result section, we will change Figure 1 as requested by transferring the back bone cartoon shown in Figure S8 into Figure 1A, citing it in the virtual screening and NMR section.

Data presentation: The authors should include an overlay of the corresponding data points (as dot plots) in Figure 2 (Bar charts), Figure S4 and Figure S5, and please, add the n number and define the error bars (e.g. SD, SEM) in the figure legends.

We improve the data representation by including of course the definition of the error bars in the figure legends of the various plots and the n numbers. Since the other two referees were ok with the presentation of the data, suggesting even that the paper is actually quite well written, we do not plan so far to change the bar charts in Fig. 2. We think they represent the data quite clearly.

Figure S1: Please use monospaced fonts to ensure that the sequences are aligned.

Figure S1 will of course be changed to monospaced fonts (courier, as it was before), and checked before submission whether the format is kept.

Equation1 Please correct this equation.

Of course, we will correct equation 1.