Scale:chr2:

20 kb69,150,000 69,160,000 69,170,000 69,180,000 69,190,000 69,200,000 69,210,000

UCSC Genes (RefSeq, UniProt, CCDS, Rfam, tRNAs & Comparative Genomics)

GKN2 GKN1GKN3

Chromosome 2 (p13.3)21 p12 34 35

MouseRat

RabbitPig

DolphinCow

HorseCatDog

MicrobatHedgehog

ElephantArmadilloOpossum

Tasmanian_devilWallaby

Multiz Alignments (Mammalian Conservation)

GKN2

Figure S1. Mammalian conservation of GKN2. (A) UCSC genome browser output for the human GKN gene cluster (GKN1-3) with expanded view of the human GKN2 structural gene on chromosome 2p13.3 (GRCh37 Hg19; http://genome.ucsc.edu/). Mammalian conservation is shownin the ‘Multiz Alignment’ track. The degree of conservation, as shown in the grayscale density plot, uses blocks of darker values to indicate higher conservation.

Mouse models:

2m 12m

Autoimmune gastritis

Tumorigenesis (antrum)

Tumorigenesis (corpus)

H. pylori gastritis 7d

Atrophy/hypertrophy

HpSS1

GmcsfTg

gp130F/F

IL1βTg

HKβ-/-

Gastric epithelium

Mucus layer

Submucosa

Helicobacter pylori

NORMAL H. PYLORI GASTRITIS

GASTRICCANCER

Tumor

Inflammatory cells

INTESTINALMETAPLASIA

Goblet cells

Human GC progression:

Figure S2. Comparison of human and mouse GC progression. Schematic showing correspondence between stages of human GC as defined by Correa (Correa P et al. Lancet. 1975. 2(7924): 58-60) and five well characterised mouse models of gastric pathology: infection of C57BL/6 with mouse adapted H. pylori Sydney strain 1 (HpSS1); granulocyte-macrophage colony stimulating factor transgenic overexpression (GmcsfTg) model of autoimmune gastritis; H+K+ ATPase beta subunit knockout (HKβ-/-) model of corpus atrophy and hypertrophy; gp130Y757F cytokine co-receptor knock-in model of antral tumorigenesis (gp130F/F); interleukin-1 beta transgenic overexpression model of corpus tumorigenesis (IL1βTg).

A

274

124

151 Gkn2-/-Gkn2+/-WT

0

2

4

6

8

10

aver

age

litte

r siz

e

Birth Weaning

Gkn2-/-WT

B

Figure S3. Gkn2-/- mice show normal viability and fertility. (A) Pie chart shows genotype frequencies of progeny resulting from Gkn2+/- heterozygous pair matings. (B) Histogramshows mean litter sizes at birth and weaning from Gkn2-/- × Gkn2-/- matings compared to WT× WT matings. Error bars (SEM).

0

1

2

3

***

WTGkn2-/-

N

umbe

r of c

orpu

s le

sion

s

A

0

2

4

6

8

M

acro

scop

ic le

sion

are

a as

% o

f gas

tric

muc

osa

area ***

B

Figure S4. Quantitative morphometric analysis of corpus mucosal lesion number and macroscopic area. (A) Number of lesion foci per corpus stomach; (B) macroscopic lesion mucosa area as % of total gastric mucosal area in 12 week old Gkn2-/- (n=17) and WT littermate control (n=9) mice. Histograms show the mean. Error bars (+SEM). P values were determined using a 2-tailed Student’st test: ***(P<0.001).

-

Tff1 Gkn2 -/- /Tff1Gkn2 -/-WT -/- -/-A

0.0

0.1

0.2

0.3

0.4

0.0

0.1

0.2

0.3

0.4B

WTGkn2-/-

Tff1-/-

Gkn2-/-/Tff1-/-

Cor

pus

area

(mm

2 )/le

ngth

mus

cula

ris m

ucos

a (m

m)

Ant

ral a

rea

(mm

2 )/le

ngth

mus

cula

ris m

ucos

a (m

m)

ns

ns

C

Figure S5. Morphometric analysis of gastric epithelial lesions in Gkn2-/-/Tff1-/- mice.(A) Representative images of fundic and antral lesions in 12-week old Gkn2-/-/Tff1-/- compound mutant mice. Boundaries between fundic and antral mucosae are delineated with a white broken line; lesion areas are demarcated with either red (fundic) or green (antral) dotted outlines. Scale bar shows 5mm. (B-C) Histograms show the mean mucosal cross section (microscopic) area of fundic (B) and antral (C) mucosae. Error bars (+SEM). Abbreviations: ns; non-significant comparison in 2-tailed Student’s t test.

-

A

0

2

4

615202530

mRN

A fo

ld c

hang

e

Dmbt1 RegIIIγ

Gkn2-/- uninfected (4)WT H. pylori (13)Gkn2-/- H. pylori (18)

B

-5

0

5

10

15

D

IL10 Foxp3

mRN

A fo

ld c

hang

e

-20

0

20

40

60

80

100

mRN

A fo

ld c

hang

e

IL17f

C

mRN

A fo

ld c

hang

e (T

H1/

M1

gene

s)

-5

0

5

10

50

100

150

Ifnγ IL1β Cxcl2 IL6 IL11Tnfα

Figure S6. Lack of differential expression of cytokines and immune-related transcription factors in antral stomach of H. pylori-infected Gkn2-/- mice. QRT-PCR analysis of anti-microbial peptides Dmbt1 and RegIIIγ (A); Th-1/M1 cytokines (B), Th-17 cytokines (C) and Treg markers (D). Histograms show the mean mRNA fold change relative to WT uninfected control mice. Error bars + SEM.

C

A

Figure S7. Enhanced gastric immunity in Gkn2-/- mice is not cell autonomous to GECs. (A) Schematic showing the strategy used for co-culture of primary mouse GEC and H. pylori SS1. (B) Immunofluorescent co-localization of epithelial cells (pan-cytokeratins; Pan-CK), GKN2 and β-gal in cultured GECs prepared from Gkn2-/- and WT mice. Absence of contaminating immunocytes and mesenchymal cells was verified by CD45 and SMA staining respectively. Primary mouse embryonic fibroblasts (F’blasts) and peritoneal macrophage (Macs) cultures show corresponding positive controls for SMA and CD45. (C) Luminex bead array analysis of cytokines/chemokines in Gkn2-/- and WT GEC culture supernatants after co-culture with H. pylori for 24 hours. Histograms show mean levels (pg/mL). Error bars (+SEM). P values were determined using a 2-tailed Student’s t test. Statistical significance relative to WT uninfected controls: *P<0.05. There were no differences in chemokine/cytokine production between H. pylori-stimulated Gkn2-/- and WT GECs.

Gkn2-/-(n=9)

WT(n=7)

Primary GECs co-cultured with 1 x 10 cfu live H. pylori SS17

Gkn2-/- GECsWT GECs

0200400600800

100010002000300040005000

1000015000200002500030000

Cyt

okin

e le

vel p

g/m

L

WT uninfectedGkn2 -/- uninfectedWT H. pyloriGkn2-/-H. pylori

TNFαIL-

1αIL-

1β IL-5

IL-6

IL-10

IL-13

CCL3CCL4

CCL5

**

**

****

**

**

**

**

**

**

GK

N2

Pan

-CK WTWT WT

WT 50µm 20µm 50µmGkn2-/- 20µm

βgal

SM

AC

D45

GEC 50µm 20µm50µm

F’blasts

Macs

Nuc

Gkn2-/-

Pan

-CK

B

0

2

4

6

*

105

H. p

ylor

i CFU

/sto

mac

h

WT Gkn2-/-

Figure S8. Confirmatory analysis of H. pylori colonisation in Gkn2-/- mice by colony forming unit (CFU) assay. Histograms show mean H. pylori CFU counts per stomach at 7 days post infection. Error bars (+SEM). P values were determined using a 2-tailed Student’s t test: *(P<0.05). The CFU assay provides independent verification of QPCR data showing lower gastric colonisation levels in Gkn2-/- compared to WT mice.

-

0.00

0.02

0.04

0.06

0.08

** * * * *

GKN2 (DOX):

J99 ∆babA/sabA

- + - +H. pylori strain :

WT26 WT31 Empty Vector

J99 ∆babA/sabA

- + - +J99 ∆babA/

sabA

- + - +

H. p

ylor

i adh

esio

n/A4

50 n

m

A

B

50 100

200

500

10000

GKN2 (HA)

GAPDH

DOX (ng/mL)

DOX (ng/mL)

50 100

200

500

10000

WT26:

WT31:

GKN2 (HA)

GAPDH

nsns

ns

Figure S9. GKN2 is dispensable for H. pylori adhesion to GECs. (A) Stable transfected, GKN2-HA (C-terminal haemagglutinin epitope tagged) inducible MKN28 gastric epithelial cell lines were created using Tet-ON Advanced/TRE-tight expression vectors. Doxycycline (DOX) dependent induction of GKN2-HA expression was verified in two indepdent clones (WT26, WT31) by immunoblotting cell lysates with anti-HA antibodies. GAPDH blots confrim protein integrity. (B) ELISA-based detection of H. pylori J99 (WT strain) and ∆babA/∆sabA (adhesion-deficient isogenic mutant strain used as control) adherence to GKN2-HA inducible MKN28 gastric epithelial cells. Histograms show mean adhesion (absorbance units at 450nm (A450). Error bars (+SEM). P values were determined using a 2-tailed Student’s t test: *(P<0.05). H. pylori J99 adhesion was unaffected by GKN2 expression (non-significant; ns).

-

Figure S10. Antral cytokine protein expression in 7 day H. pylori-infected Gkn2-/- mice. Histograms show mean levels of pro- and anti-inflammatory cytokines measured by Luminex bead array in 7 day infected mice and uninfected controls (n=6 or 7/group). Error bars (+ SEM). P values were determined using a 2-tailed Student’s t test: *P<0.05.Abbreviations: nd; not detected.

050

100150200200300400500600

pg/1

00m

g pr

otei

n

TNFα IFNγIL-1α IL-6 IL-10CXCL1 CCL4

nd nd

WT uninfectedGkn2 -/- uninfected WT H. pyloriGkn2 -/- H. pylori

*

-

B

0

400

800

1200

E-Cad CD45

Cd4

5 m

RN

A fo

ld c

hang

e

0

20

40

60

E-Cad CD45Muc

5ac

mR

NA

fold

cha

ngeWT

Gkn2-/-

0

5

10

15

20

E-Cad CD45

Cdh

1 m

RN

A fo

ld c

hang

e

0

2000

4000

6000

8000

E-Cad CD45

IL6

mR

NA

fold

cha

nge

0

5000

10000

15000

20000

Ccl

4 m

RN

A fo

ld c

hang

e

0

5000

10000

15000

20000

25000

E-Cad CD45

IL1a

mR

NA

fold

cha

nge WT

Gkn2-/-

E-Cad CD45

C

0

1000

2000

3000

E-Cad CD45

Tnfa

mR

NA

fold

cha

nge

AC

D45

E-Cadherin

WT Gkn2-/-

CD45+ CD45+

E-Cad+ E-Cad+

Figure S11. Cell sorting and cytokine expression analysis of gastric epithelial and immune cells. (A) Representative FACS plots showing isolation of gastric epithelial (E-Cadherin+) and immune cell (CD45+) populations from WT and Gkn2-/- stomachs (n=3/group). (B) QRT-PCR validation of sorted cell populations for Cdh1 (E-Cadherin) and Cd45 mRNA expression. Analysis of Muc5ac mRNA confirms presence of SMC within the gastric epithelial population. (C) QRT-PCR analysis of pro-inflammatory cytokine expression in sorted gastric epithelial and immune cell populations. IL1a, Tnfa, IL6 and Ccl4 mRNA showed specific enrichment in gastric immune cellsand were undetected in the epithelial cells. Error bars (+SEM)._

Ly6C

CD11b+ CD49d+ CD11b+ Ly6CHi

Ly6G- CD49d+

Mo-MDSC

Ly6G

CD49d

Ly6G

FSC

SS

C Leukocytes

CD

11b

F4/80

CD11b+

CD

11b

CD

11b

F4/80Ly6C

CD11b+ CD11c+

DCMacrophages

CD11b+ F4/80+

Gr1-Ly6G

FSC

SS

C

CD11c

LeukocytesGr1-

Gr1-

B

A

Figure S12. Flow cytometric gating strategy to resolve gastric mucosal myeloid cell populations. Initial gating on viable cells by propidium iodide dye exclusion (not shown). Subsequent gating was on total gastric leukocytes (forward/side scatter) for (A) macrophages, dendritic cells (DC) and (B) monocytic myeloid derived suppressor cells (Mo-MDSC) are shown.

-

WT uninfected Gkn2 -/- uninfected WT H. pylori Gkn2 -/- H. pylori

CD11b+ Ly6CHi Ly6G- CD49d+

10.9

Ly6G

Ly6C

2.7 8.7 2.8

Mo-MDSC

CD

11b

F4/80

4.6 12.1 3.9 9.5

CD11b+ F4/80+ Gr1-Macs

CD

11b

CD11c

CD11b+ CD11c+

3.4 7.0 3.2 5.7DC

Figure S13. Flow cytometric analysis of gastric mucosal immune cells in 7 day H. pylori infected Gkn2-/- mice. Representative flow plots showing key myeloid cell subsets in stomachs of 7 day infected Gkn2-/- mice and uninfected controls (n=5/group): macrophages (Macs; CD11b+F4-80+Gr1-),dendritic cells (DC; CD11c+), monocytic (Mo)-MDSC (CD11b+Ly6ChiLy6G-CD49d+).

WT Gkn2+/- Gkn2-/-A

Figure S14. Baseline gastric phenotype of Gkn2+/- mice. (A) Macroscopic images of stomachs from 12 week old mice. White broken lines delineate boundaries of corpus and antral mucosae. Scale bar 5 mm. (B) Histology of AB-PAS stained corpus and antrum mucosa in 12 week old mice. Scale bars 50 µm. (C) Morphometric analysis of corpus and antral mucosal thickness in 12 week old WT (n=8); Gkn2+/- (n=4); Gkn2-/- (n=11) mice. (D) Ki-67 labelling of proliferating cells in corpus mucosa of 12 week old WT (n=6); Gkn2+/- (n=4) and Gkn2-/- (n=6) mice. Error bars (+SEM). P values weredetermined using a 2-tailed Student’s t test: *(P<0.05). Gkn2+/- mice show a normal gastric phenotype in macroscopic appearance, mucosal histology/thickness and epithelial proliferation.

-

0.0

0.1

0.2

0.3

0.0

0.1

0.2

0.3

0.4

0.5

Muc

osal

are

a (m

m2 )/

leng

thm

uscu

laris

muc

osa

(mm

)

WT

Gkn2-/-Gkn2+/-

Corpus AntrumC

0

10

20

30

% K

i-67

posi

tive

cells

/gla

nd

WT

Gkn2 -/-Gkn2+/-

CorpusD

**

B

Cor

pus

Ant

rum

WT Gkn2+/- Gkn2-/-

A

Gkn2-/-(n=18)

WT(n=13)

(6 - 8 weeks old)

Oral delivery of 1 x 10 cfu H. pylori SS1

2

Sacrifice

0

7

Infection timecourse/months

Gkn2+/-(n=24)

Gkn2 H. pylori+/-WT H. pylori Gkn2 H. pylori-/-C

0.0

0.5

1.0

1.5

2.0

2.5

PMN MN Atrophy PMN MN

Corpus Antrum

MetPa

thol

ogy

Scor

e

DGkn2 -/- uninfected WT H. pylori

Gkn2 -/- H. pyloriGkn2+/- H. pylori

#

#

##

##

#

# # #

#

#

** * *

**

#

##

Figure S15. Gastric immunopathology in H. pylori infected Gkn2+/- mice. (A) Schematic showing strategy used for 2 month H. pylori infections and treatment group sizes. (B) QPCR analysis of H. pylori SS1 colonisation levels in 2 month infected gastric homogenates. (C) Macroscopic and corresponding histological views gastric corpus from 2 month H. pylori infected mice. Scale bars: 5 mm (macro); 50 µm (histology). (D) Summary of semi-quantitative gastric histopathology. (E) QRT-PCR analysis of TH1/proinflammatory (Ifng, IL1b, IL6, Cxcl2), anti-inflammatory (IL10), TH17 (IL17f) cytokine mRNA expression in 2 month infected gastric corpus. Error bars (+SEM). P values were determined using a 2-tailed Mann Whitney U test (Figure S12B) or a 2-tailed Student’s t test (Figure S12D-E). Statistical significance compared to WT uninfected control mice: #(P<0.05). Statistical significance between treatment groups: *(P<0.05); **(P<0.01).

-

0

0.5

1.0

1.5

104

H. p

ylor

i gen

omes

/10

5 G

apdh

cop

ies

B

*

* WT H. pylori

Gkn2 -/- H. pyloriGkn2+/- H. pylori

-5

0

5

10

15

20

25

30

Ifng

mR

NA

fold

cha

nge

-2

0

2

4

6

8

IL1b

mR

NA

fold

cha

nge

0

40

80

120

IL6

mRN

A fo

ld c

hang

e

0

2

4

6

8

IL10

mRN

A fo

ld c

hang

e

0

10

20

30

40

Cxc

l2 m

RNA

fold

cha

nge

-50

0

50

100

150

200

250

IL17

f mRN

A fo

ld c

hang

e

E#

#

#

***

# #

#*

#

##

#

##

#

###

#

##

* ***

Gkn2 -/- uninfected WT H. pylori

Gkn2 -/- H. pyloriGkn2+/- H. pylori

Table S2. Gkn2-/- mice show normal fertility. Shown are average litter sizes at

birth and weaning produced by Gkn2-/- × Gkn2-/- matings compared to WT × WT

matings. P value (Pval) significance levels from unpaired t-tests are shown. Litter

sizes resulting from Gkn2-/- matings did not differ significantly from those resulting

from WT matings.

Parents

genotype

Litters (n) Av. litter size

(at birth)

Pval Av. litter size

(at weaning)

Pval

WT 28 7.39 0.408 4.64 0.820

Gkn2-/- 44 6.86 4.44

Table S1. Gkn2 null alleles show normal Mendelian inheritance. Shown

are observed versus expected genotype frequencies for progeny produced by

matings between Gkn2+/- heterozygous pairs, chi-squared (2) statistic and P

value (Pval) significance level. Observed genotypes did not deviate

significantly from expected Mendelian ratios.

Genotype Observed

(((Expected)

% of Total 2 (deg. freedom) Pval

WT 151 (137) 27.5 2.668 (2) 0.26

Gkn2+/- 274 (275) 49.9

Gkn2-/- 124 (137) 22.6

Supplemental Methods

Loss of Gastrokine-2 drives premalignant gastric inflammation and

tumor progression

Trevelyan R. Menheniott, Louise O’Connor, Yok Teng Chionh, Jan Däbritz,

Michelle Scurr, Benjamin N. Rollo, Garrett Z. Ng, Shelley Jacobs, Angelique

Catubig, Bayzar Kurklu, Stephen Mercer, Toshinari Minamoto, David E. Ong,

Richard L. Ferrero, James G. Fox, Timothy C. Wang, Philip Sutton, Louise M.

Judd, Andrew S. Giraud

Taqman Q-PCR based determination of H. pylori colonisation level

Genomic (g)DNA was extracted from H. pylori-infected mouse stomach tissue using Trizol reagent (Invitrogen) and diluted 1:10 in 10 mM Tris-HCl buffer. Diluted gDNA was analysed by Q-PCR quantitation of H. pylori SS1 genome copies (16S rRNA gene) relative to mouse Gapdh copies as described (1) using a ABI 7500 Fast Real Time PCR analyser (fast cycling) and ABI Taqman fast reagents (Applied Biosystems) in 12 µL reactions. Cycling parameters and sequences of oligonucleotide primer pairs/fluorogenic hybridisation probes were as follows:

Initial denaturation: 95˚C for 10 mins 40 cycles: 95˚C for 3 secs 60˚C for 30 secs (fluorescence data capture)

ELISA-based detection of anti-H. pylori antibodies in mouse serum

Levels of circulating anti-H. pylori antibodies were quantified by standard direct enzyme-linked immunosorbent assay (ELISA) as described (2). Briefly, Maxisorp Immunoplates (Nunc, Roskilde, Denmark) were coated overnight with 10 μg/well of H. pylori SS1 lysate in 100 μL of bicarbonate buffer (pH 9.6). After being washed three times with tap water then one time with 0.05% (v/v) Tween 20 in PBS, wells were blocked with 1% (w/v) bovine serum albumin in PBS (PBS-BSA) for 1 hr at room temperature. Blocker was washed off as above then mouse serum samples were serially diluted 1:100 to 1:108 in PBS-BSA and 100 μL added to duplicate wells before incubation at room temperature for 30 min. After further washing, 100

Primer name Sequence (5’ fluorochrome and 3’ quencher dyes)

16S rRNA-F TTTGTTAGAGAAGATAATGACGGTATCTAA C

16S rRNA-R CATAGGATTTCACACCTGACTGACTATC

16S rRNA-Probe FAM-5'-CGTGCCAGCAGCCGCGGT-3'-TAMRA

Gapdh-F TGCACCACCAACTGCTTAG

Gapdh-R GGATGCAGGGATGATGTTC

Gapdh-Probe FAM-5'-CAGAAGACTGTGGATGGCCCT-3'-TAMRA

2

μL of either horseradish peroxidase (HRP)-conjugated goat anti-mouse immunoglobulin G1 (IgG1) diluted 1/6,000 (Southern Biotech, Birmingham, AL, USA) or HRP-conjugated goat anti-mouse IgG2c (Immunology Consultants Laboratory, Portland, OR, USA) diluted 1/5,000 in PBS-BSA was added to each well and the plates incubated at room temperature for 1 hr. Following additional washes color was developed by addition of 100 μL tetramethylbenzidine (Zymed, CA) and the reaction stopped by addition of 100 μL of 2M H2SO4. Absorbance was read at 450 nm and end point titers calculated. The observed end point titer is the highest serum dilution that yielded an optical density (OD) greater than the value that defined the cutoff between positive and negative results. The cutoff value is determined by calculating the mean of the OD of the negative control wells (PBS-BSA only) for each plate + 2 standard deviations (IgG1) or 3 standard deviations (IgG2c). Titer is expressed as the reciprocal dilution of the OD cutoff value for each sample. Where the observed values failed to reach the OD of the lowest dilution (1:100) the titer was assigned the minimum dilution value of 100.

ELISA-based determination of H. pylori adhesion to cultured human GECs

MKN28 gastric epithelial cells were cultured in 96-well plates at 37°C and 5% CO2 for 48 hrs. Growth medium was aspirated and the cells co-incubated with 107 cfu

H. pylori J99 (3) or adhesion defective H. pylori J99 babA/sabA isogenic mutant (gifts of Michael McGuckin, Mater Medical Research Institute, Brisbane, Australia; ref. (4)) in Dulbecco’s modified Eagle medium (DMEM) and the plate gently agitated for 30 min at 37°C. Cells were fixed with 1% paraformaldehyde in PBS for 10 min. Positive controls (maximal H. pylori binding to wells) comprised wells with no gastric cells, to which bacteria were added and allowed to adhere to the plastic before fixation. Negative controls contained neither gastric cells nor bacteria. After washing, cells were blocked for 30 min with PBS-BSA and bacterial adhesion determined by ELISA using hyperimmune mouse serum (generated in house) and a goat anti-mouse IgG-HRP secondary antibody (Pierce; 1/10,000) for detection. Binding was visualized by addition of 100 μL tetramethylbenzidine (TMB) substrate per well, with absorbance read at 450 nm. Generation of polyclonal antiserum Antibodies were prepared essentially as described (5). To generate anti-mouse GKN2 antibodies, a custom immunizing peptide 12-mer corresponding to the C-terminus of mouse GKN2 was synthesized (sequence: ILGVSICGGIH; molecular weight 1068 Daltons; Auspep, Australia) at >95% purity. To generate anti-human GKN2 antibodies a custom 10-mer corresponding to the C-terminus of human GKN2 was synthesized (sequence: GISICADIHV; molecular weight 1030 daltons; Auspep, Australia) at >95% purity. Mouse and human GKN2 immunising peptides were respectively coupled to key-hole limpet haemocyanin (KLH; Sigma) with 5% glutaraldehyde using 1 mg immunising peptide, 2 mg KLH and 80 µL glutaraldehyde in 1 mL PBS for 1 hr. The conjugate was dialysed overnight against PBS using 6000MW pore dialysis tubing to remove free peptide and glutaraldehyde, then frozen in aliquots for immunization. New Zealand white rabbits (two pairs; Walter and Eliza Hall Institute (WEHI) Antibody Facility, Bundoora, Victoria, Australia) were bled from a marginal ear vein (pre-immunization antibody titer) then immunized intramuscularly with 100 nM human or mouse GKN2-KLH conjugate emulsified in two volumes of complete Freund’s

3

adjuvant (Sigma). Five weeks later rabbits were boosted with 50 nM GKN2-KLH conjugate in incomplete Freund’s adjuvant (Sigma) and bled 7-10 days later (5-20 mL volumes). Two subsequent immunizations and bleeds were carried out at intervals depending on antibody titer using incomplete Freund’s adjuvant. The final bleeds (anti-mouse GKN2-R771 B3; anti-human GKN2-R779 B3) was used in all subsequent experiments at a final dilution of 1:500 (immunohistochemistry and immunofluorescence cytochemistry) and 1:3000 (immunoblotting). Co-culture of primary gastric epithelial cells with H. pylori

Primary mouse gastric epithelial cell (GEC) cultures were derived as described (6). Briefly, stomachs from 3 week old mice were dissected, opened and contents rinsed away with sterile PBS. Stomachs were transferred to 0.2% w/v bovine serum albumin (BSA) in Hanks buffered salts solution (HBSS; Sigma), chopped into ~1 mm3 pieces then digested by incubation in 0.4 mg/mL collagenase A (Roche) in HBSS at 37°C for 1 hr. Digested tissue was resuspended in DMEM Glutamax medium supplemented with 20% FBS, 2 mM non-essential amino acids, 50 IU penicillin, 50 ug/mL streptomycin (Invitrogen), disaggregated by repeated pipetting and seeded into 24-well plates. After one hr at 37°C with 5% CO2/air to allow for the attachment of contaminating fibroblasts, the unattached fraction (enriched for GECs) was aspirated, seeded in fresh 24-well plates and incubated for a further 48 hrs to allow for GEC attachment. Cells were co-cultured with 1 × 107 cfu live H. pylori SS1 for 24 hrs as described (7) after which culture supernatants were harvested for cytokine analysis. Genotyping assay for the Gkn2-(lacZ) knock-in allele

Mice were genotyped by end point PCR from ear punch biopsies using the following cycling parameters and multiplexed oligonucleotide primer sequences: Initial denaturation:

95˚C for 3 mins 35 cycles: 95˚C for 30 secs 60˚C for 1 min

72C for 1 min Final extension:

72C for 5 min

Products were analysed by electrophoresis through 1.5% agarose, Tris-borate EDTA (TBE) gels. Gkn2-(WT)-F1 and Gkn2-(WT)-R1 primers amplify a 442 bp product from the Gkn2 WT allele, Gkn2-(WT)-F1 and Gkn2-(lacZ)-R2 primers amplify a 214 bp product from the Gkn2-lacZ knock-in allele. WT, Gkn2-/- and

Primer name Sequence

Gkn2-(WT)-F1 CAGTCCTGATAAGGGAGTTTGC

Gkn2-(WT)-R1 CCCTCACAAAGTGATGAGAAGG

Gkn2-(lacZ)-R2 ATTCTCCCAATCTCTCCTCTGC

4

Gkn2+/- mice were respectively discriminated by amplification of either 442 bp, 214 bp or both 442/214 bp products (see Figure 2B, main text).

Generation of inducible human GKN2 expressing gastric epithelial cell lines

To generate MKN28 gastric epithelial cell lines with an inducible GKN2 protein expression cassette we used the Tet-ON Advanced/TRE-tight binary vector system (Clontech). A full length human GKN2 cDNA was amplified from human gastric antrum cDNA with simultaneous engineering of a C-terminal haemagglutinin (HA) epitope tag and flanking NotI-SalI restriction sites by PCR mutagenesis. Oligonucleotide primer sequences are shown in the table below with color coding used to indicate HA-tag sequence, NotI and SalI restriction sites; stop codons. Kozak consensus (bold type); GKN2 coding sequence (underlined).

Primer name Sequence

NotI-GKN2-F 5’-GCGGCCGCACCATGAAAATACTTGTGGCATTTCTGG-3’

SalI-stop/stop-HA-GKN2-R 5’-GTCGACTCACTAAGCGTAATCTGGAACATCGTATGGG- -TAAACATGAATGTCTGCACAGATTGAAATTCC-3’

Initial denaturation:

95C for 3 mins 30 cycles:

95C for 30 secs

55C for 1 min

72C for 1 min Final extension:

72C for 5 min PCR-engineered GKN2-HA cDNA product (750 bp) was purified by electrophoresis through a 1% agarose, Tris-acetate EDTA (TAE) gel, subcloned into the pGEM-T-easy vector and correct mutagenesis verified by sequencing using BigDye_v3.1 chemistry (Applied Biosystems). The GKN2-HA cDNA was excised by double digestion with NotI and SalI then directionally ligated into the NotI-SalI sites of the pTRE-tight expression vector. MKN28 cells were maintained in RPMI 1640 medium supplemented with 10% fetal

bovine serum (FBS), 50 IU/mL penicillin, 50 g/mL streptomycin (Invitrogen) at

37C with 5% v/v CO2/air (5). MKN28 cells were transfected with 2 g linearised pTet-ON Advanced vector DNA using FuGENE HD (Roche), incubated for 48 h,

then subcultured (1:10 split) and stable transgenic clones selected with 300g mL-1 G418 (Sigma) for two weeks. G418-resistant colonies were trypsinised and cloned by single cell deposition into 96-well plates using a Mo-Flo sorter (Beckman-Coulter). MKN28 Tet-ON clones with inducible potential were identified by supertransfection with pTRE-tight-luciferase in the presence and absence of 1 µg/mL doxycycline (DOX) inducer. Two clones (F6 and G6) showing robust DOX-dependent luciferase induction/repression were expanded into stable Tet-ON cell

5

lines. MKN28 Tet-ON F6/G6 cells were transfected 2 g linearised pTRE-tight-GKN2-HA as described above and double stable transgenic clones selected with

150 g/mL G418 (maintenance dose) and 200 g/L hygromycin for two weeks. G418- hygromycin double resistant colonies were single cell cloned as described above and expanded into MKN28 Tet-ON/TRE-tight-GKN2-HA continuous cell lines. Two clones (WT26 and WT31) showing appropriate DOX-dependent induction of GKN2-HA expression were used in all experiments.

Gene expression (QRT-PCR) cycling parameters and primer sequences

QRT-PCR was performed using an ABI 7500 Fast Real Time PCR analyser (Applied Biosystems) in standard cycling mode and Go-Taq SYBR-green reagents (Promega) in 20 µL reactions. Cycling parameters and oligonucleotide primer sequences used for mouse and human gene expression analysis were as follows:

Initial denaturation:

50˚C for 2 mins 95˚C for 10 mins 40 cycles: 95˚C for 15 secs 60˚C for 1 min (fluorescence data capture) Dissociation curve analysis:

60˚C 95˚C 1% increment/30secs

Human primer sequences:

Mouse primer sequences:

Gene Forward Primer Sequence Reverse Primer Sequence

Rpl32 GAGGTGCTGCTGATGTGC GGCGTTGGGATTGGTGACT

Gkn2 GAGACAGTGACCATCGACAACC GAGACAGTGACCATCGACAACC

Muc6 GCTGTGTATGACAAGTCGGGTTAC AATTTTGTCCTTCTTGGACAGATACAT

Mist1 CCAGGGTGCTCCTTCTTTTG GGCGGAAGTTCACCATCCTT

HKβ CCGGTGGGTGTGGATCAG GCAAAGAGCCCGGTCATG

Ifn CCAGGACCCATATGTAAAAGAAGC TCATGTCTTCCTTGATGGTCTCC

IL1β ATGACCTGAGCACCTTCTTTCC CTTGTTGCTCCATATCCTGTCC

Tnfα GTAGCCCACGTCGTAGCAAA ACAAGGTACAACCCATCGGC

Cxcl2 AGTGAACTGCGCTGTCAATG TTCAGGGTCAAGGCAAACTT

IL6 ACAAAGCCAGAGTCCTTCAGAGA CTGTTAGGAGAGCATTGGAAATTG

Gene Forward Primer Sequence Reverse Primer Sequence

GAPDH GACATCAAGAAGGTGGTGAAGC GTCCAACCCTGTTGCTGTAG

GKN1 GGCCTGATGTACTCAGTCAACC TTTAGTTCTCCACCGTGTCTCC

GKN2 TGCAGGATCATGCTCTTCTAC TGGTCCATCTTCAGGATAAAG

6

IL11 TCATTCAGGGAGGCTAAGGA ACAGGGTTTTGCCATGTCTC

Dmbt1 GCTCCACAAGCACAGATTCC TGCCATCTCATTGTTCTTGG

RegIII ACGAATCCTTCCTCTTCCTCAG GTCTTCACATTTGGGATCTTGC

IL10 TGATGCCCCAGGCAGAGA CACCCAGGGAATTCAAATGC

Foxp3 TTCACCTATGCCACCCTT ATCC TTTCTGAAGTAGGCGAACATGC

IL4 GTCATCCTGCTCTTCTTTCTCG TCACTCTCTGTGGTTCTTCG

IL5 TTGACAAGCAATGAGACGATGAG GCCCCTGAAAGATTTCTCCAA

Fizz1 CCTGCTGGGATGACTGCTACT CTCCACTCTGGATCTCCCAAGAT

Ym1 TCTGGTGAAGGAAATGCGTAAA GCAGCCTTGGAATGTCTTTCTC

IL1RN CATAGTGTGTTCTTGGGCATCC TCAGAGCGGATGAAGGTAAAGC

IL17a TCTGTGTCTCTGATGCTGTTGC ACATTCTGGAGGAAGTCCTTGG

IL17f AAAACCAGGGCATTTCTGTCC ACCAGGATTTCTTGCTGAATGG

Ccr2 AAGTTCAGCTGCCTGCAAAGAC GCCGTGGATGAACTGAGGTAAC

IL4ra GTGCCACATGGAAATGAATAGG TGTGAGGTTGTCTGGAGCTAGG

S100a8 GTCCTCAGTTTGTGCAGAATATAAA GCCAGAAGCTCTGCTACTCC

S100a9 CTCTTTAGCCTTGAAGAGCAAG TTCTTGCTCAGGGTGTCAGG

Arg1 AAAGCTGGTCTGTGGAAAA ACAGACCGTGGGTTCTTCAC

Arg2 ACCCTTGCGGCTCACACA AGCAGGATTGCAGACACTGAAA

Tgfb1 GGTGTGGTCTATGCCATCAA CCTCTCCAGTAACCGCTGAT

Tgfb2 CTTCACCACAAAGACAGGAACC CCATCAATACCTGCAAATCTCG

RoRt GAGTTTGCCAAGCGGCTTT TCCATTGCTCCTGCTTTCAGT

Tbet TCCAAGAGACCCAGTTCATTGC CGTATCAACAGATGCGTACATGG

Tff1 AGAGGTTGCTGTTTTGATG AGTCTGAGGGGTTGAACTG

Tff2 CCCCACAACAGAAAGAAC GGGCACTTCAAAGATCAG

Muc1 CACACTCACGGACGCTACGT TACCTGCCGAAACCTCCTCAT

Muc5ac GCAGTTGTGTCACCATCATCTGTG GGGGCAGTCTTGACTAACCCTCTT

Quantitative morphometry

Quantitative morphometry was performed by a blinded observer in all cases. At least six representative photographs per animal (n≥5) of standard histological or immunohistochemistry sections were captured using a Nikon Eclipse 80i light microscope equipped with a DS Ri1 imaging system (Nikon). To generate morphometric measurements, lengths, areas or relevant cells were manually outlined within images using ImageJ software for Windows v1.44p (http://rsb.info.nih.gov/ij/index.html). All measurements were converted from image pixels to mm (lengths) or mm2 (areas) by comparison with a calibrated graticule.

Determination of gastric acid content

For determination of gastric acid content in mice, stomachs were dissected and clamped at the oesophagus and duodenum prevent leakage of contents. Stomachs were injected with 2mL normal saline (0.9% w/v NaCl), incubated at room temperature for 5 mins with agitation, then cut open and the contents collected in

7

sterile tubes. The pH of the stomach contents was determined immediately using a pre-calibrated pH meter (3510, Jenway).

Definitions of gastric histopathology

Corpus and antral inflammation was defined by sub-mucosal and mucosal presence of polymorphonuclear and mononuclear cells. Defects in gastric epithelial differentiation were assessed by several criteria. Corpus glandular atrophy was defined as loss of parietal and zymogenic cells. Mucus neck cell (MNC) hyperplasia was defined as expansion of the neutral mucin-secreting MNC zone, without ectopic presence of acidic mucins. MNC hyperplasia is a low grade lesion that may precedes the emergence of mucus metaplasia (see below). Surface mucus cell (SMC) metaplasia describes a lineage having the morphological characteristics of SMC, but showing loss of typical neutral mucins and ectopic gain of Alcian blue-stained acidic mucins. This atypical metaplastic phenotype was frequently found in Gkn2-/- mice, particularly in association with MNC hyperplasia, but has not been widely described elsewhere. Mucus metaplasia was defined as the ectopic presence of Alcian blue stained acidic mucins associated with the acquisition of an elongated antral glandular structure within corpus glands. This metaplasia is a premalignant change, commonly observed in chronic H. pylori infections of humans and mice, as well as some mouse models of GC.

Scoring criteria for semi-quantitative gastric histopathology

Inflammation (corpus and antrum) was scored according to the extent of immune cell infiltration within the fundic mucosa: 0, no immune cells were detected in the submucosa and mucosa; 0.5, few immune cells located within the submucosa only; 1, immune infiltration of the submucosa with or without infiltration at the very base of the mucosa; 2, immune infiltration of the submucosa and the bottom half of the mucosa; 3, immune infiltration of the submucosa and greater than 50% of the mucosa; 4, transmural infiltration of immune cells. Glandular atrophy was scored based on the estimated percentage of parietal cell loss and chief cell loss within the corpus: 0, no visible parietal cell and chief cell loss; 1, 25% parietal cell loss and 50% chief cell loss; 2, 50% parietal cell loss and greater than 75% chief cell loss; 3, 75% parietal cell loss and 100% chief cell loss; 4, greater than 75% parietal cell loss and 100% chief cell loss. Mucus neck cell (MNC) hyperplasia was scored according to the approximate percentage of corpus mucosa encompassed by glandular units with an expanded MNC zone: 0, 0% of mucosa affected; 1, 25% of mucosa affected; 2, 50% of mucosa affected; 3, 75% of mucosa affected; 4, 100% of mucosa affected. Surface mucus cell (SMC) metaplasia severity was scored according to the proportion of surface mucus cells affected: 0, no visible surface mucus cell metaplasia; 1, small foci were affected; 2, up to one-third of SMC were affected; 3, one-third to two-thirds of SMC were affected; 4, greater than two-thirds of SMC were affected. Mucus metaplasia (H. pylori infected mice only) was scored based on the approximate percentage of the corpus mucosa showing replacement of oxyntic glands with elongated Alcian-blue stained glands reminiscent of antral mucosa: 0, no visible mucus metaplasia; 1, small foci were present; 2, up to one-third of the corpus was affected; 3, one to two thirds of the corpus were affected; 3, greater than two thirds of the corpus were affected.

8

Immunoblotting and multiplex cytokine array analysis

Tissue protein extracts, cell lysates and culture supernatants were size fractionated by 15% SDS-PAGE, transferred to nitrocellulose membranes (GE Healthcare) and blocked in 5% non-fat milk powder/Tris-buffered saline (TBS) pH 7.4, 0.1% Tween-

20. Membranes were incubated with primary antibodies overnight at 4C. Detection was performed with anti-rabbit/ anti-mouse IgG-HRP conjugates (DAKO) and enhanced chemiluminescence reagents (GE Healthcare). Sources of primary antibodies: rabbit polyclonal anti-mouse GKN2 (R771-B3) diluted 1:3000; and anti-human GKN2 (R779-B3) diluted 1:3000 (custom generated, see above); rabbit polyclonal anti-human GAPDH (Abcam, #ab9485) diluted 1:3000. Cytokine protein content of mouse stomach tissues and cell culture supernatants was determined using Bio-Plex Pro Mouse 23-plex cytokine bead arrays (Bio-Rad) according to the manufacturer’s protocols. In brief, stomach tissues were homogenized in ice cold BHI supplemented with complete mini protease inhibitor cocktail (Roche), homogenates cleared by centrifugation, supernatants collected and diluted 1:4 in BHI prior to analysis. Cell culture supernatants were centrifuged to remove cell debris. A nine point standard curve was generated by serial dilution of normalized cytokine standard in diluents specific to each assay (i.e. BHI broth or RPMI/DMEM culture medium) to ensure that the matrix used for assay calibration matched that of the samples. Samples were added in 50µL volumes either diluted 1:4 (stomach homogenates) or undiluted (culture supernatants). Plates were scanned on a Luminex 200 analyzer using xPONENT software (Luminex Corporation). Whole mount staining for β-galactosidase in situ

Stomach tissue expression of the lacZ reporter gene was assessed by histochemical staining for encoded β-galactosidase activity as described (8). Stained tissues were post fixed in 4% paraformaldehyde (Sigma) in PBS overnight at room temperature and then processed for standard paraffin histology as described (9). Supplemental Methods References

1. Roussel Y, Harris A, Lee MH, and Wilks M. Novel methods of quantitative real-

time PCR data analysis in a murine Helicobacter pylori vaccine model. Vaccine.

2007;25(15):2919-29.

2. Chionh YT, Ng GZ, Ong L, Arulmuruganar A, Stent A, Saeed MA, Wee JL, and

Sutton P. Protease-activated receptor 1 suppresses Helicobacter pylori gastritis via

the inhibition of macrophage cytokine secretion and interferon regulatory factor 5.

Mucosal immunology. 2015;8(1):68-79.

3. Alm RA, Ling LS, Moir DT, King BL, Brown ED, Doig PC, Smith DR, Noonan B,

Guild BC, deJonge BL, et al. Genomic-sequence comparison of two unrelated

isolates of the human gastric pathogen Helicobacter pylori. Nature.

1999;397(6715):176-80.

4. Mahdavi J, Sonden B, Hurtig M, Olfat FO, Forsberg L, Roche N, Angstrom J,

Larsson T, Teneberg S, Karlsson KA, et al. Helicobacter pylori SabA adhesin in

persistent infection and chronic inflammation. Science. 2002;297(5581):573-8.

5. Menheniott TR, Peterson AJ, O'Connor L, Lee KS, Kalantzis A, Kondova I,

Bontrop RE, Bell KM, and Giraud AS. A novel gastrokine, Gkn3, marks gastric

atrophy and shows evidence of adaptive gene loss in humans. Gastroenterology.

2010;138(5):1823-35.

9

6. Viala J, Chaput C, Boneca IG, Cardona A, Girardin SE, Moran AP, Athman R,

Memet S, Huerre MR, Coyle AJ, et al. Nod1 responds to peptidoglycan delivered

by the Helicobacter pylori cag pathogenicity island. Nat Immunol. 2004;5(11):1166-

74.

7. Wee JL, Chionh YT, Ng GZ, Harbour SN, Allison C, Pagel CN, Mackie EJ,

Mitchell HM, Ferrero RL, and Sutton P. Protease-activated receptor-1 down-

regulates the murine inflammatory and humoral response to Helicobacter pylori.

Gastroenterology. 2010;138(2):573-82.

8. Charalambous M, Menheniott TR, Bennett WR, Kelly SM, Dell G, Dandolo L, and

Ward A. An enhancer element at the Igf2/H19 locus drives gene expression in both

imprinted and non-imprinted tissues. Dev Biol. 2004;271(2):488-97.

9. Howlett M, Chalinor HV, Buzzelli JN, Nguyen N, van Driel IR, Bell KM, Fox JG,

Dimitriadis E, Menheniott TR, Giraud AS, et al. IL-11 is a parietal cell cytokine that

induces atrophic gastritis. Gut. 2012;61(10):1398-409.