Supporting Information Rivron et al. 10.1073/pnas.1117627109 SI Text SI Materials and Methods. Isolation and culture of hMSCs. Bone mar- row aspirates were obtained from donors with written informed consent and hMSCs were isolated and proliferated as described previously (1). Alternatively, hMSC were obtained from Lonza (Lonza Group Ltd.) (mentioned as “Cambrex” cells in the text). Briefly, aspirates were resuspended by using 20-gauge needles, plated at a density of 5 × 10 5 cells per square centimeter and cul- tured in hMSC proliferation medium containing α-MEM (Life Technologies), 10% FBS (Cambrex), 0.2 mM ascorbic acid (Asap; Life Technologies), 2mM L-glutamine (Life Technologies), 100 units∕mL penicillin (Life Technologies), 10 μg∕mL strepto- mycin (Life Technologies), and 1 ng∕mL basic FGF (Instruchem- ie). Cells were grown at 37 °C in a humid atmosphere with 5% CO 2 . Medium was refreshed twice a week, and cells were used for further subculturing or cryopreservation. The coculture differentiation medium was composed of Dulbecco’s Modified Eagle’s Medium, 10 −7 M dexamethasone, 50 mg∕mL ascorbate 2-phosphate, 40 mg∕mL proline, 100 mg∕mL pyruvate, and 50 mg∕mL ITS 1 Premix (Becton—Dickinson, MA: 6.25 mg∕ mL insulin, 6.25 mg∕mL transferrin, 6.25 ng∕mL selenious acid, 1.25 mg∕mL bovine serum albumin, 5.35 mg∕mL linoleic acid). Human umbilical vein endothelial cells. Human umbilical vein en- dothelial cells (huvECs) were purchased from Lonza (Lonza Group Ltd.). Cells were grown at 37 °C in a humid atmosphere with 5% carbon dioxide (CO 2 ) in endothelial growth medium-2 (Lonza Group Ltd.). Cells were routinely split at a 1∶5 ratio. Cells from passage 3 or 4 were used for coculture experiments. Coculture in a modified hanging-drop system. Coculture of hMSC (92%) and hUVEC (8%) were obtained by resuspending a total of 150,000 cells per spherical aggregate in 2 mL of differentiation medium and then seeded in a well of a Deepwell 96 well-plate (Nunc). A meniscus was formed by slightly overfilling the wells. The plate was then inverted to form a drop hanging from each well. Cells accumulated by gravity at the tip of the hanging drop (air-liquid interface) and formed a spherical aggregate within 48 h. Plates were placed inverted, using plastic spacers, on a lid and cultured at 37 °C in a humid atmosphere with 5% CO 2 . Pic- tures of spherical aggregates were taken using an inverted micro- scope. Medium was changed every third day by inverting the plate and transferring the spherical aggregates to a new plate. Gene expression analysis by q-PCR and microarray. The effect of cy- clopamine on gene expression was analyzed by seeding a cocul- ture of 92% hMSCs and 8% huvEC (150,000 cells in total) in Deepwell 96 well-plates supplemented or not in cyclopamine, for 12 d. Medium was changed every third day by transferring the spherical aggregate to a new plate. Three pools of 12 aggre- gates each were formed for each condition and RNA was isolated by using an RNeasy mini kit (Qiagen). To study the genome-wide effect of cyclopamine, gene expression profiling was performed using arrays of HumanU133_2-type from Affymetrix. Biotiny- lated antisense cRNA was then prepared according to the Affy- metrix standard labeling protocol. Afterwards, the hybridization on the chip was performed on a GeneChip Hybridization oven 640, then dyed in the GeneChip Fluidics Station 450 and there- after scanned with a GeneChip Scanner 3000. All of the equip- ment used were from the Affymetrix-Company (Affymetrix). A Custom CDF Version 12 with Entrez based gene definitions was used to annotate the arrays (2). The raw fluorescence inten- sity values were normalized applying quantile normalization. Dif- ferential gene expression was analyzed based on ANOVA using a commercial software package SAS JMP7 Genomics, version 4, from SAS (SAS Institute). A false positive rate of a ¼ 0.05 with FDR correction was taken as the level of significance. Repre- sented genes are all significantly regulated (n ¼ 3) and the values are log2-converted fold-changes. The effect of Shh on mRNA expression was analyzed by seed- ing a coculture of 92% hMSCs and 8% huvEC (150,000 cells in total) in Deepwell 96 well-plates supplemented or not with Sonic Hedgehog, from day 4 to 12. Medium was changed every third day by transferring the spherical aggregate to a new plate. RNA was isolated by using an RNeasy mini kit (Qiagen), and qPCR was performed by using SYBR green (Invitrogen) and established pri- mers (SABioscience) on a MyiQ2 detection system (Biorad). Data were analyzed using the Biorad software, using the fit point method by setting the noise band to the exponential phase of the reaction to exclude background fluorescence. Expression of chondrogenic marker genes was calculated relative to GAPDH levels by the comparative CT method. Histological analysis. After harvesting, spherical aggregates were frozen in Cryomatrix at −60 °C. Sections (7 μm) were cut with a cryotome. Sections were fixed in cold acetone for 5 min and air-dried, rehydrated for 10 min, after which they were incubated for 30 min with 10% FBS in PBS to block nonspecific background staining. Sections were incubated with mouse antihuman CD31 (Dako) (dilution 1∕20) or Phalloidin-Alexa fluor 488 conjugated antibody (Invitrogen) (dilution 1∕30) for 1 h. Sections were washed in PBS and subsequently incubated with the secondary antibody (Alexa Fluor 494 antibody, Invitrogen). Samples were counterstained with Dapi (Sigma). Tissue explants (5 weeks) were fixed in paraformaldehyde, decalcified overnight in 0.1 mM EDTA and embedded in paraffin. Planar sections (6 μm) were cut with a microtome. The decalcification of the tis- sues results in a softer tissue and allows for sectioning in the main plan. Masson-Goldner (Merck Chemicals) and TRACP (Takara) staining kits were used according to the manufacturer’s instruc- tions. CD31 (dilution 1∕80) and SMA (1∕80) (Dako) were incu- bated for 1 h and subsequently incubated with the secondary antibody (horseradish peroxidase conjugated goat-anti-mouse immunoglobulin antibody, Dako) for 45 min. Slides were devel- oped with diaminobenzidine (Dako) as substrate and were weakly counterstained with hematoxylin (Sigma). For Fig. 5 A–D, tissue explants (8 weeks) were directly embedded in methyl methacry- late (Sigma) for sectioning. Approximately 10 μm-thick, undecal- cified sections were processed on a histological diamond saw (Leica saw microtome cutting system). For Fig. 5 E–H, tissue ex- plants (8 weeks) were embedded into paraffin. Undecalcified cross-sections were processed on a microtome (7 μm-thick). The sections were stained with basic fuchsin and methylene blue to visualize new bone formation. The newly formed mineralized bone stains red with basic fuchsin, all other cellular tissues stain light blue with methylene blue, and the ceramic material remains black and unstained by both dyes. Histological sections were qua- litatively analyzed using a light microscope (Leica), and quanti- tative histomorphometry was performed using high-resolution digital photographs (300 dpi) of the middle section of three total sections from each implant. Before histomorphometrical analysis, bone was pseudocolored using Photoshop CS2 (Adobe Systems) and normalized to the total surface area of the implant or to the total surface of the ceramic scaffold. Rivron et al. www.pnas.org/cgi/doi/10.1073/pnas.1117627109 1 of 4