Glucagon Ucn3 DAPIGlucagon Ucn3

Somatostatin Ucn3 DAPISomatostatin Ucn3

PPY Ucn3 DAPIPPY Ucn3

A

HG

FED

CB

I

Supplementary Figure 1: Ucn3 expression in mouse islets is restricted to β-cells.(A-C) Confocal images showing immunostaining of Ucn3 (green) and glucagon (red) on adult mouse pancreatic sections. (D-F) Ucn3 (green) and somatostatin (red) (G-I) Ucn3 (green) and pancreatic polypeptide (PPY, red). Nuclei are stained with DAPI (blue). Scale bars = 50μm. No co-localization of Ucn3 is seen with any of the islet hormones (other than insulin – see Fig 2 and 3).N

atur

e B

iote

chno

logy

: doi

:10.

1038

/nbt

.214

1

Ucn3 Insulin

E18.5

P6

P13

E18.5

P6

P13

(2713±805)

(2062±93)

(1907±189)

(1234±55)

(2675±86)

(5365±380)

Signal intensity

Supplementary Figure 2: Ucn3 expression levels increase gradually in all β-cells during maturation, whereas insulin content stays constant.Intra-cellular FACS analysis of insulin and Ucn3 in E18.5 (blue), P6 (green) and P13 (red). Histograms present the signal intensity of Ucn3 (A) and insulin (B) plotted against the percentage of all insulin expressing cells. Numbers in brackets show mean intensity ± sem of three independent biological repeats (three separate litters) for each age group.

Nature Biotechnology: doi:10.1038/nbt.2141

Glucagon UCN3 DAPIGlucagon UCN3

D E

Somatostatin UCN3 DAPISomatostatin UCN3

IHG

PPY UCN DAPIPPY UCN3

KJ L

Insulin UCN3

A CB

F

Insulin UCN3 DAPI

Supplementary Figure 3: UCN3 expression in human pancreas(A-C) Confocal images showing immunostaining of UCN3 (green) and insulin (red) on pancreatic sections from an adult human. (D-F) UCN3 (green) and glucagon (red). (G-I) UCN3 (green) and somatostatin (red) (J-L) UCN3 (green) and pancreatic polypeptide (PPY, red). Nuclei are stained with DAPI (blue). Scale bars = 50μm. Nature Biotechnology: doi:10.1038/nbt.2141

Supplementary Figure 4: hESC-derived β-cells secrete human C-peptide in response to glucose challengeMice transplanted with 5 million hESC-derived pancreatic progenitors were fasted 12 weeks after transplantation over night and injected with 2mg/kg glucose. The levels of human C-peptide before (fasting, blue) and one hour after glucose administration (glucose, red) are shown. Despite variable basal levels of human C-peptide, all animals except mouse #4, showed glucose-stimulated secretion of human C-peptide.

Nature Biotechnology: doi:10.1038/nbt.2141

Glucagon UCN3 DAPIGlucagon UCN3

D FE

Somatostatin UCN3Somatostatin UCN3

IHG

PPY UCN3 DAPIPPY UCN3

KJ L

Insulin UCN3 DAPIInsulin UCN3

A CB

Supplementary Figure 5: UCN3 expression in hESC-derived β-cells after maturation in vivo(A-C) Confocal images showing immunostaining of UCN3 (green) and insulin (red) on hESC-derived graft 8 months post transplantation. (D-F) UCN3 (green) and glucagon (red). (G-I) UCN3 (green) and somatostatin (red) (J-L) UCN3 (green) and pancreatic polypeptide (PPY, red). Nuclei are stained with DAPI (blue). Scale bars = 50μm. Nature Biotechnology: doi:10.1038/nbt.2141

Supplementary table 1 Up regulated genes

PROBE_ID Gene symbol E18.5 P1 P10 Adult Fold

change ILMN_2673260 Ctrl* 49 ± 8 69 ± 30 2269 ± 2057 3110 ± 1910 59.1 ILMN_2881083 Try10* 65 ± 29 300 ± 312 4572 ± 3844 7237 ± 4442 39.7 ILMN_2716989 Prss2* 46 ± 6 81 ± 54 1663 ± 1192 2392 ± 1334 39.4 ILMN_3160208 Cpb1* 48 ± 17 83 ± 51 1627 ± 1322 2169 ± 1747 34.7 ILMN_2693403 Ela1* 88 ± 26 102 ± 33 1513 ± 758 5071 ± 2642 34.0 ILMN_2666677 Cel* 60 ± 23 95 ± 48 2100 ± 1965 1871 ± 1213 33.4 ILMN_2493756 Try4* 49 ± 14 156 ± 141 1985 ± 1574 3706 ± 2554 33.1 ILMN_2728429 1810010M01Rik* 53 ± 14 90 ± 52 1051 ± 632 3239 ± 2226 31.5 ILMN_2592415 Reg1* 106 ± 71 202 ± 141 4954 ± 3212 3272 ± 2018 30.1 ILMN_2671137 Ela3b* 57 ± 27 125 ± 94 1752 ± 1573 2280 ± 1351 26.7 ILMN_1220763 Rnase1* 69 ± 38 128 ± 90 1891 ± 1290 3065 ± 2175 26.7 ILMN_2670847 Cpa1* 89 ± 38 107 ± 52 1718 ± 1287 2728 ± 1855 24.6 ILMN_3104915 1810049H19Rik* 47 ± 12 141 ± 122 1493 ± 1201 1953 ± 1200 23.3 ILMN_2919377 Ctrb1* 106 ± 59 225 ± 203 3360 ± 2265 2670 ± 2042 21.2 ILMN_2933478 Amy2-2* 47 ± 9 83 ± 43 428 ± 228 2166 ± 1554 20.2 ILMN_1246265 Clps* 233 ± 182 467 ± 419 6237 ± 3526 6551 ± 4551 18.5 ILMN_2963762 Try10l* 50 ± 15 163 ± 155 1272 ± 1049 1887 ± 1138 18.5 ILMN_2874291 Amy2* 81 ± 45 205 ± 186 1080 ± 588 4393 ± 2661 18.3 ILMN_1216509 Ctrc* 40 ± 4 43 ± 8 274 ± 196 1084 ± 920 17.9 ILMN_2990661 Pnliprp2* 38 ± 5 47 ± 13 502 ± 355 697 ± 568 17.3 ILMN_2722659 Clps* 86 ± 43 149 ± 113 1829 ± 1128 1668 ± 1335 16.7 ILMN_2674620 Ela2* 253 ± 286 627 ± 576 5169 ± 3159 10910 ± 5518 16.5 ILMN_1232533 Sycn* 46 ± 12 72 ± 34 462 ± 292 1374 ± 1022 16.5 ILMN_2829699 EG436523* 51 ± 24 167 ± 160 1127 ± 957 1320 ± 996 14.1 ILMN_2708477 Spink3* 52 ± 13 98 ± 76 993 ± 641 529 ± 383 13.3 ILMN_2760199 Klk6 57 ± 13 74 ± 23 564 ± 384 829 ± 609 12.3 ILMN_2904435 Gp2* 37 ± 1 41 ± 7 181 ± 80 622 ± 493 11.2 ILMN_1225909 Pnliprp1* 451 ± 380 774 ± 664 8347 ± 1724 4806 ± 2403 9.3

1

Nature Biotechnology: doi:10.1038/nbt.2141

Up regulated genes (cont.)

PROBE_ID Gene symbol E18.5 P1 P10 Adult Fold

change ILMN_2844820 Angptl7 68 ± 4 105 ± 16 383 ± 125 1043 ± 151 8.8 ILMN_1216566 LOC232680* 61 ± 9 64 ± 13 452 ± 198 447 ± 349 8.2 ILMN_2731191 Klk5* 46 ± 4 49 ± 10 238 ± 157 343 ± 269 7.5 ILMN_2684115 2210010C04Rik* 37 ± 2 48 ± 13 201 ± 147 297 ± 184 7.4 ILMN_1238143 Ucn3 127 ± 20 154 ± 13 923 ± 22 1234 ± 285 7.2 ILMN_2690014 Syt4 80 ± 4 62 ± 5 252 ± 22 734 ± 107 6.9 ILMN_2692167 Pnliprp2* 40 ± 2 43 ± 7 176 ± 130 264 ± 205 6.7 ILMN_1226556 2310032F03Rik* 44 ± 3 47 ± 4 172 ± 15 435 ± 207 6.6 ILMN_2860932 Zbtb2* 91 ± 13 76 ± 9 348 ± 211 629 ± 600 6.6 ILMN_1259215 Serpina10 136 ± 10 97 ± 25 275 ± 18 1303 ± 13 6.6 ILMN_1252131 Klk1b27 44 ± 5 52 ± 13 193 ± 151 270 ± 216 6.1 ILMN_1228211 Tff2 95 ± 3 81 ± 17 409 ± 215 446 ± 334 6.0 ILMN_1238736 Klk1b4 50 ± 3 58 ± 15 211 ± 141 308 ± 225 5.9 ILMN_1231724 Resp18 577 ± 54 684 ± 113 2693 ± 221 4401 ± 528 5.6 ILMN_2850077 Adh1 159 ± 21 195 ± 18 654 ± 25 1232 ± 221 5.1 ILMN_2824971 Gpr158* 149 ± 7 163 ± 25 617 ± 88 900 ± 134 5.0 ILMN_2592718 Cuzd1* 39 ± 4 39 ± 2 104 ± 58 235 ± 167 4.9 ILMN_1258501 Adh1 141 ± 23 161 ± 7 555 ± 19 999 ± 242 4.8 ILMN_2871660 Car15 52 ± 4 55 ± 1 364 ± 35 120 ± 45 4.7 ILMN_2968692 Cpa2 41 ± 5 44 ± 3 164 ± 80 152 ± 94 4.4 ILMN_2960700 Prf1 277 ± 43 249 ± 43 821 ± 302 1281 ± 918 4.2 ILMN_2804685 Defb1 69 ± 6 133 ± 42 319 ± 19 536 ± 126 4.2 ILMN_2596522 Mt1* 958 ± 521 1348 ± 109 2719 ± 1045 7849 ± 4383 4.1 ILMN_2822825 Fbxo2 90 ± 1 111 ± 24 314 ± 50 401 ± 27 3.8 ILMN_2695819 Ddit4l 49 ± 4 50 ± 4 168 ± 51 149 ± 46 3.6 ILMN_2728038 Arhgap24 139 ± 16 128 ± 28 348 ± 33 595 ± 215 3.4 ILMN_2648742 Abcb4 78 ± 20 90 ± 19 266 ± 14 362 ± 46 3.4 ILMN_2839027 Tceal6 95 ± 9 102 ± 21 238 ± 10 447 ± 41 3.4 ILMN_2994299 Hgfac 182 ± 43 243 ± 18 547 ± 46 965 ± 91 3.3

2

Nature Biotechnology: doi:10.1038/nbt.2141

Up regulated genes (cont.)

PROBE_ID Gene symbol E18.5 P1 P10 Adult Fold

change ILMN_2628647 Ddc 560 ± 97 553 ± 76 1408 ± 133 2434 ± 256 3.3 ILMN_3064283 Pde4dip 115 ± 3 93 ± 17 245 ± 7 433 ± 45 3.3 ILMN_2681232 D12Ertd647e 216 ± 19 218 ± 20 492 ± 31 931 ± 203 3.2 ILMN_1254335 Rgpr 50 ± 3 47 ± 4 112 ± 13 194 ± 27 3.2 ILMN_3108770 Fbxl10 117 ± 18 116 ± 19 309 ± 1 489 ± 51 3.2 ILMN_1250689 Rgs9 133 ± 27 107 ± 14 361 ± 22 456 ± 49 3.1 ILMN_2624854 Gstm2 146 ± 23 140 ± 28 316 ± 36 603 ± 61 3.1 ILMN_1251449 Gstm2 144 ± 21 126 ± 11 303 ± 16 578 ± 64 3.1 ILMN_2601471 Ccnd1 177 ± 64 196 ± 13 473 ± 92 763 ± 165 3.0 ILMN_2959272 Rnu6 669 ± 62 511 ± 128 1627 ± 739 1383 ± 630 3.0 ILMN_3162403 St6galnac3 70 ± 4 79 ± 13 175 ± 3 280 ± 14 3.0 ILMN_2647234 Dio1 62 ± 9 65 ± 7 135 ± 13 248 ± 98 2.9 ILMN_3125606 D12Ertd647e 215 ± 22 214 ± 2 449 ± 29 800 ± 78 2.8 ILMN_2856567 Ppy 1597 ± 704 2056 ± 957 5681 ± 1998 4387 ± 3417 2.8 ILMN_2966162 Tmem56 88 ± 8 74 ± 20 188 ± 17 266 ± 91 2.8 ILMN_1217118 Enpp5 280 ± 46 281 ± 19 587 ± 2 1071 ± 343 2.7 ILMN_1221503 Ccnd1 206 ± 81 207 ± 21 466 ± 97 788 ± 134 2.7 ILMN_2862470 Gstm2 81 ± 11 77 ± 12 163 ± 0 270 ± 21 2.6 ILMN_2646640 1700019D03Rik 84 ± 7 91 ± 18 185 ± 2 278 ± 53 2.6 ILMN_2615096 Dpp4 119 ± 20 106 ± 17 312 ± 0 306 ± 19 2.5 ILMN_2652757 Elovl5 501 ± 20 510 ± 89 1153 ± 120 1247 ± 103 2.4 ILMN_2722996 Ptpns1 407 ± 77 424 ± 89 883 ± 13 1240 ± 203 2.4 ILMN_2731769 Plekhb2 149 ± 20 161 ± 24 341 ± 35 365 ± 60 2.2 ILMN_2652414 Ncald 146 ± 13 141 ± 23 312 ± 12 337 ± 31 2.2

3

Nature Biotechnology: doi:10.1038/nbt.2141

Down regulated genes

PROBE_ID Gene symbol E18.5 P1 P10 Adult Fold

change ILMN_1244618 Dlk1 2028 ± 195 2051 ± 315 126 ± 18 39 ± 1 -23.3 ILMN_2946520 Npy 3464 ± 545 2001 ± 551 235 ± 25 77 ± 38 -17.8 ILMN_2643658 Ghrl 1199 ± 617 643 ± 205 77 ± 27 40 ± 4 -17.1 ILMN_2755578 Nnat 1677 ± 588 1286 ± 499 257 ± 79 69 ± 17 -8.8 ILMN_2649773 Slc38a5 7037 ± 586 5761 ± 660 1181 ± 68 421 ± 122 -8.0 ILMN_1232456 Nnat 593 ± 146 448 ± 102 98 ± 28 42 ± 2 -7.1 ILMN_2598022 Ghrl 337 ± 147 173 ± 38 50 ± 2 43 ± 2 -6.9 ILMN_2518412 Grb10 1459 ± 160 933 ± 133 246 ± 11 155 ± 50 -6.2 ILMN_2635700 Lgi2 823 ± 136 833 ± 113 131 ± 10 166 ± 63 -5.8 ILMN_1251414 Nxf 525 ± 159 510 ± 141 89 ± 27 94 ± 48 -5.7 ILMN_2643049 Chst8 419 ± 65 611 ± 109 113 ± 16 64 ± 7 -5.7 ILMN_2842601 Gp9 356 ± 64 240 ± 49 62 ± 15 48 ± 3 -5.3 ILMN_1215713 Egr4 514 ± 250 434 ± 157 130 ± 51 46 ± 4 -5.3 ILMN_2597769 Igf2 1412 ± 843 852 ± 433 302 ± 106 232 ± 92 -4.9 ILMN_2906728 H19 239 ± 186 122 ± 107 54 ± 18 41 ± 2 -4.8 ILMN_2629519 Cryba2 4385 ± 749 4145 ± 949 1153 ± 153 624 ± 50 -4.8 ILMN_2623983 Egr2 247 ± 110 186 ± 96 63 ± 10 45 ± 5 -4.6 ILMN_2597827 Arc 321 ± 147 288 ± 91 90 ± 29 46 ± 8 -4.6 ILMN_2619408 Atf3 198 ± 110 115 ± 55 47 ± 6 43 ± 4 -4.4 ILMN_2734712 Ptpla 551 ± 81 372 ± 48 139 ± 5 87 ± 13 -4.3 ILMN_2708203 Cdkn1c 262 ± 67 203 ± 34 67 ± 2 55 ± 3 -4.3 ILMN_1250438 Mlp 233 ± 41 190 ± 64 59 ± 7 47 ± 1 -4.1 ILMN_2592834 Sct 274 ± 35 177 ± 23 68 ± 3 49 ± 3 -4.0 ILMN_1218913 Igf2bp3 307 ± 30 254 ± 45 94 ± 15 39 ± 1 -4.0 ILMN_2834379 Tgfbi 175 ± 30 155 ± 8 47 ± 5 41 ± 4 -3.9 ILMN_2745551 Olfml2b 228 ± 23 133 ± 38 45 ± 6 48 ± 5 -3.8 ILMN_2687661 Mfng 385 ± 25 427 ± 76 157 ± 29 41 ± 9 -3.7 ILMN_2945030 Col9a2 258 ± 18 218 ± 30 82 ± 4 52 ± 2 -3.6

4

Nature Biotechnology: doi:10.1038/nbt.2141

Down regulated genes (cont.)

PROBE_ID Gene symbol E18.5 P1 P10 Adult Fold

change ILMN_2981542 Mfap2 189 ± 53 171 ± 50 59 ± 4 51 ± 8 -3.6 ILMN_2771237 Lbp 161 ± 119 251 ± 167 70 ± 4 74 ± 45 -3.6 ILMN_1226175 Igf2bp3 255 ± 16 187 ± 25 85 ± 4 39 ± 4 -3.6 ILMN_2718330 Cish 1136 ± 211 1235 ± 112 421 ± 78 226 ± 92 -3.6 ILMN_2747543 Actn3 192 ± 23 217 ± 37 66 ± 4 52 ± 4 -3.5 ILMN_1221750 Lmyc1 193 ± 18 216 ± 22 73 ± 9 39 ± 1 -3.5 ILMN_2900653 Gadd45b 396 ± 125 306 ± 38 105 ± 50 87 ± 18 -3.4 ILMN_2636403 Axud1 248 ± 155 169 ± 23 84 ± 11 75 ± 19 -3.4 ILMN_1244169 Sftpd 244 ± 32 258 ± 55 87 ± 6 66 ± 8 -3.4 ILMN_1222084 Rem2 564 ± 60 634 ± 197 229 ± 47 112 ± 47 -3.2 ILMN_1225764 1700018O18Rik 133 ± 18 150 ± 11 47 ± 4 43 ± 5 -3.2 ILMN_2828112 Igfbpl1 134 ± 12 106 ± 8 38 ± 2 38 ± 2 -3.2 ILMN_2776034 Gal 202 ± 18 185 ± 20 48 ± 5 75 ± 19 -3.2 ILMN_2445324 Zfpm1 426 ± 58 489 ± 80 205 ± 3 105 ± 22 -3.1 ILMN_2941790 Cldn6 433 ± 86 344 ± 39 148 ± 9 123 ± 45 -3.1 ILMN_2736478 Doc2b 187 ± 25 134 ± 52 63 ± 3 47 ± 2 -3.1 ILMN_2930602 Doc2b 176 ± 39 114 ± 54 57 ± 7 44 ± 4 -3.0 ILMN_1246201 Cacna1h 218 ± 44 141 ± 54 62 ± 3 68 ± 21 -3.0 ILMN_2514292 Zyx 355 ± 86 264 ± 64 108 ± 16 112 ± 20 -3.0 ILMN_2678714 Idb4 195 ± 28 195 ± 30 84 ± 14 43 ± 1 -3.0 ILMN_2638324 Gnas 105 ± 34 121 ± 13 45 ± 4 40 ± 6 -2.9 ILMN_2700608 Stx1a 310 ± 45 341 ± 16 138 ± 12 93 ± 15 -2.9 ILMN_2718217 2310057H16Rik 248 ± 58 195 ± 36 91 ± 3 80 ± 5 -2.9 ILMN_2715546 Gpx3 3096 ± 199 3125 ± 456 1018 ± 207 1034 ± 229 -2.8 ILMN_2731407 Gdf3 131 ± 23 108 ± 15 44 ± 1 48 ± 6 -2.8 ILMN_1240857 Cox7a1 149 ± 26 128 ± 28 58 ± 3 46 ± 5 -2.8 ILMN_2714031 1300002F13Rik 716 ± 184 576 ± 164 278 ± 3 246 ± 42 -2.8 ILMN_1256702 S100a10 574 ± 73 455 ± 94 157 ± 36 200 ± 72 -2.8 ILMN_2756704 9130213B05Rik 195 ± 32 161 ± 27 78 ± 1 61 ± 4 -2.8

5

Nature Biotechnology: doi:10.1038/nbt.2141

Down regulated genes (cont.)

PROBE_ID Gene symbol E18.5 P1 P10 Adult Fold

change ILMN_2728379 Ivd 852 ± 99 1010 ± 106 368 ± 11 325 ± 51 -2.8 ILMN_2604029 Klf2 252 ± 98 187 ± 63 89 ± 42 66 ± 16 -2.7 ILMN_2632206 Gnas 1494 ± 455 1575 ± 250 708 ± 41 609 ± 135 -2.6 ILMN_2731949 Copeb 273 ± 86 238 ± 81 108 ± 33 91 ± 33 -2.6 ILMN_2430220 Tmem2 174 ± 34 164 ± 16 77 ± 5 65 ± 6 -2.5 ILMN_1248740 Sema3f 138 ± 31 123 ± 25 60 ± 7 49 ± 6 -2.5 ILMN_2766651 Mafb 115 ± 11 104 ± 33 47 ± 3 41 ± 2 -2.5 ILMN_3008110 Actn3 136 ± 15 141 ± 22 64 ± 2 57 ± 5 -2.4 ILMN_2735184 Col18a1 259 ± 79 210 ± 61 92 ± 56 86 ± 6 -2.4 ILMN_2996683 Pvrl2 376 ± 21 320 ± 43 159 ± 16 134 ± 34 -2.3 ILMN_2697361 B930096L08Rik 121 ± 16 120 ± 21 60 ± 6 47 ± 5 -2.3 ILMN_2745614 1810015C04Rik 278 ± 37 248 ± 94 122 ± 4 122 ± 33 -2.3 ILMN_3133448 Mfge8 412 ± 138 409 ± 52 165 ± 87 195 ± 77 -2.1 ILMN_2646625 Jun 456 ± 106 460 ± 158 220 ± 74 196 ± 90 -2.1 ILMN_2696299 D5Ertd579e 509 ± 27 523 ± 128 245 ± 50 245 ± 51 -2.0

Supplementary Table 1

List of genes differentially expressed between immature (E18.5 and P1) and mature (P10 and adult) β-cells. Numbers represent normalized expression levels in the microarray ± standard deviation. Asterisks mark acinar-related genes. Underlined genes were chosen for analysis at the protein levels.

6

Nature Biotechnology: doi:10.1038/nbt.2141

Materials and methods Animal experiments and islet isolation All animal experiments were performed in compliance with the Harvard University International Animal Care and Use Committee (IACUC) guidelines. Mouse strains used in this study were ICR, Swiss-Webster wild type and Pdx1–GFP transgenic and SCID-beige mice. Blood glucose levels were measured using OneTouch Ultra2 glucometer (LifeScan). Blood insulin levels were measured with an Ultrasensitive Insulin ELISA kit (Alpco). For glucose tolerance test, animals were fasted over-night and blood was taken form tail tips before, and 1 hour after, injection of 2gr/kg body weight glucose. Human C-peptide levels were measured using Human C-peptide ELISA kit (Alpco). For islet isolation, adult pancreata were perfused through the common bile duct with 0.8mM Collagenase P (Roche) and fetal and neonatal pancreata were dissected wholly without perfusion. Pancreata were digested with 0.8mM Collagenase P (Roche) and purified by centrifugation in Histopaque gradient (Sigma). Glucose stimulated insulin secretion (GSIS) assays Isolated islets were recovered over night in islet media (DMEM containing 1gr/L glucose, 10% v/v FBS, 0.1% v/v Penicillin/Streptomycin). Islets were picked manually under a fluorescent dissecting microscope according to their GFP fluorescence. Care was taken to pick islets of approximately the same size from all ages. For dynamic GSIS, approximately 50 islets were hand picked and assayed on a fully automated Perifusion System (BioRep). Chambers were sequentially perfused with 0.5mM, 2.8mM or 16.7mM glucose in KRB buffer (128mM NaCl, 5mM KCl, 2.7mM CaCl2, 1.2mM MgCl2, 1mM Na2HPO4, 1.2mM KH2PO4, 5mM NaHCO3 10 HEPES, 0.1% BSA) at a flow rate of 0.1ml/min. Fractions were collected and kept at -80°C until analysis. For static GSIS assays, approximately 10 islets were hand picked, incubated for 2 hours in KRB buffer at 37°C, 5%CO2, and then incubated for 75min with 2.8mM or 16.7mM glucose in the same conditions. Insulin concentrations in the supernatant were determined using Ultrasensitive Insulin ELISA kit (Alpco). Analysis of the results was done using Matlab software. Electron microscopy Samples were fixed with 4% paraformaldehyde and 0.1% glutaraldehyde for 2 h at room temperature and further refixed with a mixture of 1% osmiumtetroxide (OsO4) plus 1.5% potassium ferrocyanide (KFeCN6) for 2 h, washed in water and stained in 1% aqueous

uranyl acetate for 1 h followed by dehydration in grades of alcohol (50%, 70%, 95%,

2 × 100%) and propyleneoxide (1 h). Samples were then infiltrated in propyleneoxide:Epon 1:1 overnight and embedded in TAAB Epon (Marivac Canada Inc.). Ultrathin sections (about 60–80 nm) were cut on a Reichert Ultracut-S microtome, picked up on to copper grids, stained with 0.2% lead citrate and examined in a Tecnai G2 Spirit BioTWIN transmission electron microscope. Images were taken with an AMT CCD camera. The number of insulin vesicles and cell area were determined using ImageJ software.

1

Nature Biotechnology: doi:10.1038/nbt.2141

Microarray analysis Islets were isolated as above from heterozygous Pdx1–GFP (crossed with ICR) animals and further dissociated into single cells with 0. 25% Trypsin-EDTA (Invitrogen). GFP+ cells were isolated using FACSaria (BD Bioscience). Total RNA was extracted with RNeasy RNA extraction kit (Qiagen). Biotinylated cRNA was prepared from ≥ 100 ng of isolated RNA using Illumina TotalPrep RNA Amplification Kit (Ambion) and hybridized to the Illumina mouse genome Bead Chips (MouseRef8). All samples were prepared as four biological replicates. Data were acquired with Illumina Beadstation 500 and were evaluated using BeadStudio Data Analysis Software (Illumina). Immunohistochemistry and FACS analyses For immunohistochemistry, pancreata were fixed by immersion in 4% paraformaldehyde overnight at 4 °C. Samples were washed with PBS, incubated in 30% sucrose solution overnight and embedded with optimal cutting temperature compound (Tissue-Tek). 10μm sections were blocked with 10% donkey serum (Jackson Immunoresearch) in PBS/0.1% Triton X and incubated with primary antibodies overnight at 4°C. Secondary antibodies were incubated for 1 hr at room temperature. The following primary antibodies and dilutions were used: rabbit anti-mouse or anti-human Ucn3 (1:600-1:800, both from Phoenix Pharmaceuticals), rabbit anti-human Ucn3 (1:600, a gift from Dr. Wylie Vale, Salk Institute), Guinea Pig anti-insulin (1:800, DAKO), Guinea Pig anti-glucagon (1:200, Linco), Goat anti-Somatostatin (1:200, Santa Cruz) and Goat anti-PPY (1:200, Novus). Secondary antibodies were: Alexa Fluor 488 donkey anti-rabbit (1:400, Invitrogen), Alexa Fluor 647 donkey anti- goat (1:400, Invitrogen) and DyLight 649 donkey anti-guinea pig (1:400, Jackson Immunoresearch). Nuclei were visualized with DAPI. It is important to note that, in our hands, anti-human immunostaining was successful only on human tissues fixed over-night with 4% paraformaldehyde directly after surgery. Efforts to use the abovementioned anti-human Ucn3 antibodies on flash-frozen-unfixed cryosections, acetone-fixed cryosections or formalin-fixed-paraffin-embedded samples resulted in either close-to-background or non-specific staining. Images were taken using an Olympus IX51 Microscope or Zeiss LSC 700 confocal microscope. For intra-cellular FACS analysis, islets were isolated as above and further dissociated into single cells with 0. 25% Trypsin-EDTA (Invitrogen). The cells were then fixed with Cytofix/Cytoprem solution (BD Biosciences) at 4°C for 30min, washed once with Perm Wash Buffer (BD Biosciences), and stained with Guinea Pig anti-insulin (1:800, DAKO) and rabbit anti-Ucn3 (1:600, Phoenix Pharmaceuticals) for 1 hour at room temperature. The cell were then washed once with Perm Wash Buffer (BD Biosciences), incubated with TexasRed donkey anti-guinea pig (1:400, Jackson Immunoresearch) and Alexa Fluor 488 donkey anti-rabbit (1:400, Invitrogen) for 45 min at room temperature, washed with PBS, filtered through a nylon mash, and analyzed LSR-II FACS machine (BD Biosciences). Analysis of the results was done using FlowJo software. HESC Culture and Differentiation Human ESCs (WA1) were cultured on Matrigel (BD Biosciences) in mouse embryonic fibroblast conditioned media (MEF-CM). MEF-CM media was produced by conditioning

2

Nature Biotechnology: doi:10.1038/nbt.2141

media for 24 days on a confluent layer of mouse embryonic fibroblasts and subsequently adding 20 ng/ml bFGF (Invitrogen). The media was composed of DMEM/F12 (GIBCO) media supplemented with 20% KnockOut Serum Replacement (GIBCO), 2mM L-glutamine (L-Glu, GIBCO), 1.1mM 2-mercaptoethanol (GIBCO), 1mM nonessential amino acids (GIBCO), 1x penicillin/streptomycin (GIBCO). Cells were passaged at the ratio of 1:6–1:20 every 4-7 days using TrypLE Express (Invitroge). To initiate differentiation, the cells were cultured as previously described1 onto 1:30 dilution of growth factor reduced matrigel (BD Biosciences) in MEF-CM. Two to three days following seeding the differentiation was initiated as follows: cells were exposed to RPMI 1640 (Invitrogen) supplemented with 0.2% fetal bovine serum (FBS) (Hyclone, Utah), 100 ng/mL activin-A (AA; Pepro-tech; Rocky Hill, NJ), and 20 ng/mL of Wnt3A (R&D Systems) for day one only. For days 2-3, cells were cultured in RPMI with 0.5% FBS and 100 ng/mL AA (stage 1). During days 4-5 cells were treated with DMEM-F12 medium containing 2% FBS and 50ng/ml FGF7 (Peprotech) (stage 2). For days 6-9 cells were treated with DMEM-HG (Invitrogen), 1% (v/v) B27 (Invitrogen), 2uM RA (Sigma), 0.25uM SANT-1 (Sigma), and 100ng/ml rhNoggin (R&D Systems) (stage 3). During days 10-13 cells were treated with DMEM-HG (Invitrogen) + 1% (v/v) B27 (Invitrogen), 100ng/ml rhNoggin (R&D Systems), 50nM TPB (PKC activator, EMD Biosciences), and 1 µM ALK5 inhibitor II (Axxora, San Diego, CA) (stage 4). On day 14, cells were treated with 5 mg/mL Dispase for 5 min, followed by gentle pipetting to mix and break the cell clumps into small clusters (< 100 micron). The cell clusters were cultured for one day in a 125 ml Spinner Flask (Corning) at 80-100 rpm overnight with DMEM-HG supplemented with 1 µM ALK5 inhibitor II, 100 ng/mL of Noggin and 1% B27. For transplantation into mice, 10 million cells in clusters were transplanted under the kidney capsule of SCID-Bg mice (Jackson Laboratory). 8 months following transplant the graft was surgically extracted from under the mouse kidney capsule, fixed in 4% paraformaldehyde (PFA, Sigma), equilibrated in 30% sucrose, embedded in O.C.T., cryopreserved and sectioned. 1. Rezania, A.(2012). Maturation of human embryonic stem cell-derived pancreatic

progenitors into functional islets in diabetic mice. Submitted.

3

Nature Biotechnology: doi:10.1038/nbt.2141