Islet Interleukin-1bImmunoreactivity Is an EarlyFeature of Cystic Fibrosis ThatMay Contribute to b-Cell FailureDiabetes Care 2018;41:823–830 | https://doi.org/10.2337/dc17-1387

OBJECTIVE

Cystic fibrosis–related diabetes (CFRD) is a common complication of cystic fibrosis(CF), increasing patient morbidity and mortality. Poor understanding of CFRD path-ogenesis limits the development of targeted therapies to treat and/or prevent thedisease. The aim of this study was to evaluate islet pathology, specifically, inflam-mation, amyloid deposition, and endocrine cell composition in subjects with CF withdiabetes and with CF without diabetes.

RESEARCH DESIGN AND METHODS

A retrospective analysis of archived pancreas tissue collected at autopsy was con-ducted using pancreas tissue from subjects with CF and diabetes (CFRD) (n = 18) andCFwithout diabetes (CF-noDM) (n = 17). Two cohorts of control non-CF subjectswereidentified, each matched to CFRD and CF-no DM subjects for age, sex, and BMI(non-CF older, n = 20, and non-CF younger, n = 20), respectively. Immunohistochem-istry was performed to assess interleukin-1b (IL-1b) and islet hormone (insulin,glucagon, somatostatin, and pancreatic polypeptide) immunoreactivity; histochem-istry was performed to quantify amyloid deposition.

RESULTS

Islet IL-1b immunoreactivity was substantially increased in both CFRD and CF-no DMsubjects compared with non-CF subjects and was common in young subjects with CF(£10 years of age). In contrast, islet amyloid deposition was increased only in CFRDsubjects. We also observe abnormal islet hormone immunoreactivity, characterizedby increased glucagon immunoreactivity, in CF-no DM and CFRD subjects comparedwith non-CF subjects.

CONCLUSIONS

These findings reveal novel molecular pathways and therapeutic targets underlyingislet pathology in CF subjects andmay be important in developing newapproaches totreat CFRD.

Cystic fibrosis–related diabetes (CFRD) occurs in up to 50% of adults with cystic fibrosis(CF), with the highest incidence occurring during the second to third decade of life (1),although abnormal glucose tolerance is common even in very young children with thedisease (2–4). CFRD is associated with increasedmorbidity andmortality in bothmalesand females (5). Insulin therapy is the current standard of care for CFRD. Unlike type 1diabetes, CFRD does not appear to be the result of islet autoimmunity; HLA

1Department of Medicine, University of Wash-ington, Seattle, WA2Department of Pediatrics, University of Wash-ington, Seattle, WA3Department of Pathology, University of Wash-ington, Seattle, WA4Department of Comparative Medicine, Univer-sity of Washington, Seattle, WA5Department of Pediatrics, University of Califor-nia, San Francisco, San Francisco, CA6Diabetes Center, University of California, SanFrancisco, San Francisco, CA

Corresponding author: Rebecca L. Hull, rhull@uw.edu, or Srinath Sanda, srinath.sanda@ucsf.edu.

Received 11 July 2017 and accepted 6 January2018.

This article contains Supplementary Data onlineat http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc17-1387/-/DC1.

© 2018 by the American Diabetes Association.Readers may use this article as long as the workis properly cited, the use is educational and notfor profit, and the work is not altered. More infor-mation is available at http://www.diabetesjournals.org/content/license.

Rebecca L. Hull,1 Ronald L. Gibson,2

Sharon McNamara,2 Gail H. Deutsch,3

Corinne L. Fligner,3 Charles W. Frevert,1,4

Bonnie W. Ramsey,2 and Srinath Sanda5,6

Diabetes Care Volume 41, April 2018 823

PATH

OPHYSIO

LOGY/COMPLIC

ATIO

NS

susceptibility alleles are not observedwith greater frequency, although studiesdiffer with respect to the prevalence ofislet autoantibodies (6,7). Physiologically,CFRD appears to have more in commonwith type 2 diabetes. Genetic susceptibil-ity loci for type 2 diabetes appear to con-fer risk for CFRD (8–10), and, similar totype 2 diabetes, CFRD is characterizedby a mixture of peripheral, particularlyhepatic, insulin resistance and significantdefects in b-cell function (4,11–13). How-ever, the cause of b-cell dysfunction inCFRD is not known.Work in various animal models, includ-

ing the recently developed pig and ferretmodels, demonstrates abnormalities inglucose metabolism and insulin release,mostly in the newborn period (14–16).Human autopsy studies have been infor-mative in describing islet cell histologicalchanges with clinical manifestations. Somesmall studies have reported that subjectswithCFRDhavea lowerdensityofendocrine(islet) tissuecomparedwithpatientswith CFwithoutdiabetes (17) and/or loss ofb-cellswith no change in glucagon-producinga-cells or somatostatin-producing d-cellscompared with CF subjects without di-abetes (18,19). However, one largerstudy found no difference in b-cell areaamong subjects with CF with or withoutdiabetes (20).As in type 2 diabetes, pancreatic islet

amyloid deposition has been identifiedas a characteristic feature of CFRD in hu-mans (19,20). Islet amyloid is a well-described pathological feature of type 2diabetes, and consistent with the etiologyof this form of diabetes, islet amyloid de-position is usually observed in older indi-viduals (21–23). Islet amyloid depositionis toxic to b-cells, correlating with bothreduced b-cell area and increased b-cellapoptosis in patients with type 2 diabetes(21–23). Recent data have linked islet am-yloid deposition to islet inflammation,namely, showing that islet amyloid indu-ces inflammatory cytokine production,chiefly interleukin-1b (IL-1b), by macro-phages and dendritic cells (24,25). In asimilar fashion, the systemic inflamma-tory process observed in CF subjectsalso seems to be skewed toward IL-1b–regulated cytokines such as IL-8 and IL-17(26–28).Given these previous observations of

islet amyloidosis in CFRD (19,20), the in-flammatory properties of islet amyloid(24,25), and the systemic inflammatory

milieu of CF (26–28), we hypothesizedthat CFRD would be characterized notonly by islet amyloid but also by islet ex-pression of the inflammatory cytokineIL-1b. We conducted a histologic studyto test this hypothesis. We observedgreater IL-1b immunoreactivity in isletsof subjects with CFwith orwithout diabe-tes compared with control subjects with-out CF, while islet amyloid depositionwasonly increased in those individuals diag-nosed with CFRD. Islet endocrine compo-sition was also disturbed, with increaseda-cell area observed in subjects with CFwith and without diabetes. Based onthese observations, we propose that CFis characterized by islet inflammation,which could predispose to b-cell failure,and by islet a-cell expansion, while isletamyloid formation is mainly restricted toCFRD.

RESEARCH DESIGN AND METHODS

SubjectsForty-one patients with CF were identi-fiedby retrospective screening of autopsyrecords at the Seattle Children’s Hospital(SCH) and the University of WashingtonMedical Center for a CF diagnosis. Wescreened records from SCH between1977 and 2012 and at University of Wash-ington Medical Center between 1991 and2013. Available clinical data (Table 1), in-cluding history of lung transplantation,were extracted from the medical recordand managed using REDCap electronicdata capture tools hosted at the Univer-sity of Washington’s Institute of Transla-tional Health Sciences. Three additionalcases of CF were available from theJDRF-supported Network for PancreaticOrgan Donors with Diabetes (nPOD).Nine were excluded from further analysisbased on the following criteria: incom-pletemedical records (n = 1), lack of avail-ability or poor quality of pancreas tissue(n = 5), absence of pancreatic islets on theavailable pancreas sections (n = 2), or his-tory of islet antigen 2 antibodies and zinctransporter 8 autoantibodies, suggestiveof the presence of type 1 diabetes (n = 1).Of the remaining 35 subjects with CF,18 were classified as “CFRD” based onpresence of diabetes diagnosis by thetreating physician and/or insulin use inthemedical record; the other 17were des-ignated “CF-no DM.”Given the differencein age and BMI between the CFRD andCF-no DM cohorts, two groups of non-CF

control subjectswithoutdiabetes (including3 from the JDRF nPOD registry) were iden-tified. These, designated “non-CF (youn-ger)” (n = 20) and “non-CF (older)” (n =21), were matched for age, sex, and BMIto CF-no DM and CFRD subjects, respec-tively, and to the extent possible, giventhat autopsies were performed overmany years, these were also matched tothe time of autopsy. One non-CF controlsubject [from the non-CF cohort (older)]was excluded from further analysis, as nopancreas specimenwas available, bringingthe sample size from the non-CF (older)group to 20. The study was approved byinstitutional review boards at the Univer-sity of Washington and SCH.

Pancreatic tissue was obtained duringautopsies performed at the University ofWashington and SCHor via the nPODpro-gram. Histologic sampling from the bodyof the pancreas was routinely performed,although autopsy records did not alwaysindicate the specific location of the sample.Specimens were included in the study onlyif they showed no or minimal autolysis, asassessed by CLF and RLH (University ofWashington samples) or GHD and RLH(SCH samples). Pancreatic weight was notavailable; therefore, endocrine cell data arepresentedas relativearea rather thanmass.

ImmunohistochemistryFour-micron-thick sections of formalin-fixed, paraffin-embedded pancreas weresubjected to immunohistochemistry (IHC)(LeicaBondMax; LeicaMicrosystems, Buf-falo Grove, IL) as follows. The followingprimary antisera were used, for each ofwhich the host species is denoted insquare brackets: insulin (A0564 [guineapig], 1:4,000; Dako, Carpenteria, CA),IL-1b (49-4960 [rabbit], 1:500; ProSci In-corporated, Poway, CA; and ab2105 [rab-bit], 1:1,000; Abcam, Cambridge, MA),IL-1Ra (AF-280-NA [goat], 1:500; R&D Sys-tems, Minneapolis, MN; and NBP1-32568[rabbit], 2mg/mL; Novus Biologicals, Little-ton, CO), CD68 (Clone 514H12 [mouse], 0.34mg/mL; Leica Biosystems/Novacastra),glucagon (EP3070 [rabbit], 1:10,000;Epitomics, Burlingame, CA), somatostatin(SC-7819 [goat], 1:200; Santa Cruz Bio-technology, Santa Cruz, CA), and pancre-atic polypeptide (PP) (NB100-1793 [goat],1:250; Novus). Isotype-matched irrele-vant IgG-negative controls were used forall antisera (except those against islet hor-mones, which have been previously vali-dated in pancreas).

824 Islet IL-1b Immunoreactivity in Cystic Fibrosis Diabetes Care Volume 41, April 2018

Sections were deparaffinized followedby antigen retrieval (citrate buffer at100°C for 20min for IL-1b [both antisera];EDTA at 100°C for 10–20 min for IL-1Ra[both antisera], CD68, and PP; and pro-teinase K for 15 min at 37°C for somato-statin). No antigen retrieval was used forinsulin or glucagon IHC. Sections then un-derwent peroxide block followed by incu-bation in 10% (v/v) normal goat serum for20 min at room temperature. Sectionswere then incubated in primary antisera.For primary antisera raised in mouse,goat, or guinea pig, linking IgGs (LeicaPost Primary, rabbit anti-goat IgG, or rab-bit anti-guinea pig IgG, respectively) werethen applied (since the Leica polymer re-agent recognizes rabbit IgG). Antibodybindingwas detected by poly–horseradishperoxidase polymerized secondary detec-tion and Leica Bond Mixed Refine 3,39-diaminobenzidene detection reagents(bothDS9800; Leica Biosystems) followedby hematoxylin-eosin counterstaining andcoverslipping.Whole pancreas sections were then

digitized (Nanozoomer Digital Pathologysystem; Hamamatsu Corporation, Bridge-water, NJ). Islets (an average of 53 persection) were classified as IL-1b positiveor negative by manual counting. This wasdone by three independent observers,blinded to subject grouping (R.L.H., S.S.,

and Megan Larmore, a staff memberfrom the University of Washington CysticFibrosis Research and Translation Center’sHost Response Core). For insulin, gluca-gon, somatostatin, and PP IHC, total pan-creas tissue and hormone-positive areaswere determined automatically based onpixel value and density (Visiopharm soft-ware, Hoersholm, Denmark) and verifiedbymanual examination of segmented im-ages, as we have done previously (29).Endocrine cell areas were expressed aspercentage of total pancreatic tissuearea. Individuals performing IHC andquantitative analysis thereof were blindedto the group assignment of specimens.

Additionally, sections underwent im-munofluorescence labeling for insulin(I2018, clone K36AC10 [mouse] at 1:2,000;Sigma-Aldrich, St. Louis, MO) followed byCy3-conjugated goat anti-mouse IgGs (di-luted 1:200; Jackson ImmunoResearch,West Grove, PA) to aid the visualization ofislets and counterstaining with thioflavinS (0.5%v/v) to visualize amyloiddeposition.Islets were classified as thioflavin S positiveor negative by manual counting of at least50 islets per section, from which the pro-portion of amyloid-positive islets was de-termined for all subjects. Again, individualsperforming (immuno)histochemistry andquantitative analysis thereof were blindedto the group assignment of specimens.

Data and Statistical AnalysisComparisons between groups were doneusing one-way ANOVA for overall statisti-cal differences; pairwise comparisonswere conducted with a Tukey test to cor-rect P values for multiple comparisons. AFisher exact test was used to comparefrequency of lung transplantation be-tween groups, and a x2 test was used tocompare frequency of islet IL-1b andamyloid positivity among groups. In com-parison of differences between lungtransplant groups, two-way ANOVA wasused to test for interactions betweentransplant and diabetes status on IL-1band amyloid deposition. Linear regressionwas used to determine correlations be-tween insulin and glucagon areas.

RESULTS

Clinical DataSubjects in the CFRD group were on aver-age older than in the CF-no DM group(Table 1); however, by design the non-CF(older) and non-CF (younger) subjectgroups were matched for age and BMIto the CFRD and CF-no DM groups, re-spectively. Pancreatic insufficiency wasdocumented in all but three CF case sub-jects (for whom data were not available),regardless of diabetes status. Subjectswith CFRD had diabetes for ;5 years,

Table 1—Subject characteristics, transplant history, diabetes medications, and causes of death

Non-CF (younger) Non-CF (older) CF-no DM CFRD

N 20 20 17 18

Age (years) 13.1 6 6.8 28.0 6 5.5*# 13.9 6 8.3 28.6 6 9.0*#

Sex (female/male) 12/8 9/11 12/5 10/8

BMI (kg/m2) 19.8 6 7.0 22.6 6 5.5# 16.5 6 3.9 21.7 6 3.3#

Diabetes duration (years) (n = 17) 4.80 6 5.16

HbA1c (%) (n = 11) 7.25 6 1.67

Lung transplant (yes/no/unknown) 0/19/0 0/20/0 4/13/0 10/6/2

Other transplant 1 (heart) 0 0 0

Diabetes medicationsOral agent 1Insulin 16Unknown 2

Steroid use (yes/no/unknown) 4/13/3 0/15/5 4/13/0 11/5/2

Pancreatic enzyme supplementation (yes/no/unknown) 16/0/1 16/0/2

Cause of deathRespiratory 5 5 14 14Cardiovascular 4 4 2 d

CNS 3 6 d dInfection 4 3 d d

Malignancy 2 d d d

Multiorgan failure 1 2 d 3Unknown 1 d d 1

Data are means6 SD or n. CNS, central nervous system. *P, 0.05 vs. non-CF (younger). #P, 0.05 vs. CF-no DM.

care.diabetesjournals.org Hull and Associates 825

and all but two were insulin-treated. A his-tory of lung transplantation was more com-mon in the CFRD compared with CF-no DMgroup, and steroid use closely correlatedwiththisparameter;onlyonenon-CFsubjecthad received organ (heart) transplantation.

Islet IL-1b Immunoreactivity Is aCommon Feature of CFWe first determined whether there wasan increased IL-1b immunoreactivity inthe islets of subjects with CF (ProSci anti-body) (Fig. 1). Eighty-nine percent ofCFRD subjects exhibited some degreeof islet IL-1b positivity, with 57% of theislets on average from each subject beinginvolved (Fig. 1M). Further, islet IL-1bwasalso observed in 76% of CF-no DM casesubjects (P = 0.02 vs. CFRD), and this IL-1b

positivity was observed in 29% of isletsin this group. In contrast, only 10% ofnon-CF (younger) subjects exhibited anyIL-1b–positive islets (P, 0.001 vs. CF-noDM), comprising only 6% of islets per sub-ject. No IL-1b immunoreactivity was ob-served in the islets of non-CF (older)subjects (Fig. 1M). Validation of isletIL-1b immunoreactivity was performedusing a second anti–IL-1b antibody (Ab-cam) in a subset of subjects (n = 10 [rep-resentative examples are shown inSupplementary Fig. 1]). These subjectswere from both CF and non-CF groups andincluded subjects in whom IL-1b had beenpreviously found to be positive or nega-tive. The presence or absence of IL-1bwasconfirmed in 9 out of 10 subjects. The lastsubject was a non-CF (younger) subject

who had been classified as IL-1b positivewith the ProSci antibody but showed noas IL-1b immunoreactivity with the Ab-cam antibody. We believe the discrepancyin thisfinding isdueto lower sensitivityof thelatter antibody.

Islet IL-1b Immunoreactivity Is SeenEven in Very Young Subjects With CFWe next analyzed islet IL-1b immunore-activity (using the ProSci antibody) in thesubset of pediatric CF subjects ,10 yearsof age, the currently recommended age tobegin screening CF patients for CFRD withoral glucose tolerance tests (5). All six sub-jects with CF in this age range were in theCF-no DM group. Five of these showed ev-idenceof IL-1b immunoreactivity affecting24% of islets per subject and thus were re-flective of the CF-no DM group as a whole.

Islet Amyloid Occurs Predominantlyin CFRD SubjectsIslet amyloid was present in 61% of CFRDsubjects, involving 20% of islets (Fig. 1N),while islet amyloid was present in 18% ofCF-no DM subjects (P , 0.001 vs. CFRD),with only 3% of islets affected. Islet amy-loid was not observed in either group ofnon-CF control subjects (Fig. 1N). Of note,none of the six young CF-no DM sub-jects ,10 years of age described aboveexhibited amyloid deposition.

Effect of Prior Lung Transplantation onIslet IL-1b and Amyloid DepositionGiven the use of lung transplantation inadvanced CF disease management, CF-noDM and CFRD subjects were subdividedbasedon lung transplant statusand thepres-ence of IL-1b and amyloid was analyzed.Therewas no statistically significant inter-action between transplant status and di-abetes status in relation to islet IL-1bimmunoreactivity or amyloid deposition(Table 2). Of note, a history of steroiduse closelymirrored that of lung transplan-tation. Accordingly, subdivision of subjectsbased on steroid usage did not reveal dif-ferences in islet IL-1b immunoreactivity oramyloid deposition (data not shown).

CF Is Not Characterized by Islet IL-1Raor CD68 ImmunoreactivityWe next sought to determine whetherdifferences in the balance of IL-1b andits antagonist, IL-1Ra, exist between sub-jects with CF with and without diabetes.Based on data from older subjects withand without type 2 diabetes (30), the ex-pectation is that high IL-1Ra levels would

Figure 1—IHC for insulin, IL-1b, and amyloid in subjects with and without CF. Representativestaining of pancreas specimens from subjects in non-CF control (A–D), CF-no DM (E–H), andCFRD (I–L) groups, identifying islet b-cells using IHC for insulin (brown in A, E, and I and red in D,H, and L), IL-1b (brown; images shown from two subjects per group inB and C, F andG, and J and K),and amyloid (visualized by thioflavin S [ThioS] histochemistry; green in D, H, and L). Quantitation ofthe proportion of islets positive for IL-1b (M) and amyloid (N). Islet IL-1b positivity was increased inCF-no DM and CFRD subjects compared with age-matched non-CF control subjects, whereas isletamyloid deposition was only increased in CFRD subjects. Scale bar = 50 mm.

826 Islet IL-1b Immunoreactivity in Cystic Fibrosis Diabetes Care Volume 41, April 2018

be present in CF-no DM subjects, withlower levels being seen in CFRD subjects.Contrary to this expectation, however, lit-tle to no IL-1Ra immunoreactivity was ob-served in islets from any group despitethe use of two well-validated antibodies.With the first antiserum (AF-280-NA),IL-1Ra staining was indistinguishablefrom that seen with an isotype-matchedirrelevant IgG (data not shown). With thesecond (NBP1-32568), only sparse isletstaining was detected, with no differencein IL-1Ra immunoreactivity being observedbetweensubjectswithandsubjectswithoutdiabetes (Supplementary Fig. 2).Pancreas sections were also stained for

CD68 to investigate whether macrophageinfiltration could explain the presence ofislet IL-1b.While CD68-positive cellswereclearly present in exocrine pancreas ofsubjects with CF-no DM and CFRD, no in-filtration of macrophages was observedwithin islets, regardless of IL-1b positivityin all cases but one (Supplementary Fig. 3),suggesting that the observed islet IL-1bimmunoreactivity did not derive frommac-rophages. Only in one case, an infant (in theCF-no DM group), was intra-islet CD68 im-munoreactivity shown (Supplementary Fig.3E), suggesting that islet macrophagesmay be present during early stages of CF.

CF and CFRD Are Characterized byAbnormal Islet Hormone CellCompositionWe next analyzed islet endocrine cellcomposition in non-CF (younger andolder), CF-no DM, and CFRD subjects.

When normalized to the total pancreatictissue area, there was no significant dif-ference in insulin-positive area betweenthe groups (Fig. 2A). Insulin-positive areawas also computed as a proportion ofislet area in a subset of subjects (n = 5–10per group), yielding the same result(b-cell area/islet area was 39 6 0.6% inCF-no DM, 276 0.2% in CFRD, and 3960.4% in non-CF control subject; P = 0.13for the comparison of CF-no DM withCFRD and P = 0.95 for CF-no DM vs.non-CF control). However, a significant in-crease in glucagon immunoreactivity wasobserved in CF-no DM and CFRD subjectscompared with non-CF control subjects(Fig. 2B). A significant correlation existedbetween insulin- and glucagon-positiveareas in non-CF subjects (younger andolder cohorts) and in CF-no DM subjectsbut not in CFRD subjects (Fig. 2C–F). In asubset of patients, pancreas sections werealso stained for somatostatin andPP.Whileno significant difference between groupswasobserved for somatostatin immunoreac-tivity, CF-no DM subjects exhibited statis-tically greater pancreatic polypeptidearea compared with CFRD subjects. (Fig.2G and H). As with IL-1b and amyloid, wedetected no significant interaction betweentransplant status (or steroid use) and diabe-tes status in predicting islet hormone immu-noreactivity in CF subjects (Table 2 and datanot shown).

CONCLUSIONS

In this study,we identified islet inflamma-tion, detected by IL-1b immunoreactivity,

inmore thanhalf of subjectswith CFwith-out diabetes, and in the vast majority ofsubjects with CFRD, whereas islet IL-1bimmunoreactivity was essentially absentfrom age-matched subjectswithout CF. Inline with previous studies (19,20), isletamyloid was observed in subjects withCFRD. However, islet IL-1b was far morefrequently observed than islet amyloid inCF subjects, both those with and thosewithout diabetes, and therefore likelydoes not occur as a consequence of isletamyloid deposition. This increase in isletIL-1b in CFRD versus CF-no DM couldhave occurred as a result of the increasedage (and possibly age-related changessuch as fibrosis) in the former cohortrather than the presence/absence of di-abetes. Importantly, however, we ob-served islet IL-1b immunoreactivity inpediatric subjects ,10 years of age, thecurrently recommended age to beginscreening CF patients for CFRD with oralglucose tolerance tests (5). The preva-lence of islet IL-1b immunoreactivity inthis young cohort was similar to that inthe CF-no DM group as a whole, sug-gesting that islet inflammation couldbegin very early in CF patients, consis-tent with clinical studies showing de-rangements in glucose tolerance andb-cell function in very young subjectswith CF (2–4). Given that IL-1b is knownto contribute to impaired islet functionand viability (31), increased productionof IL-1b in the islets of CF subjects maycontribute to impaired b-cell functionthat characterizes CFRD.

Table 2—Subject characteristics and islet morphometric analyses in individuals with CF, subdivided according to diabetes andlung transplant status

CF-no DM CFRD

No lung transplant Lung transplant No lung transplant Lung transplant

N 13 4 6 10

Age (years) 11.9 6 2.4 20.3 6 1.0 22.8 6 3.4 31.4 6 2.8

Sex (female/male) 11/2 1/3 3/3 6/4

BMI (kg/m2) 15.2 6 0.6 20.3 6 2.8 21.9 6 1.3 22.0 6 1.2

Diabetes duration (years) d d 6.6 6 3.1 4.2 6 1.0

HbA1c (%) (n = 4–6) d d 7.8 6 0.9 6.5 6 0.4

Subjects with islet IL-1b, n (%) 11 (91) 2 (50) 6 (100) 8 (80)

IL-1b–positive islets (%) 28.8 6 8.5 31.1 6 18.8 63.5 6 9.5 61.8 6 12.3

Subjects with islet amyloid, n (%) 2 (11) 1 (25) 3 (50) 7 (70)

Amyloid-positive islets (%) 3.3 6 3.2 0.3 6 0.3 5.2 6 3.2 31.0 6 12.2

Insulin-positive area (%) 1.5 6 0.3 2.8 6 0.9 1.1 6 0.4 1.3 6 0.2

Glucagon-positive area (%) 1.2 6 0.3 3.5 6 0.6 1.5 6 0.5 2.5 6 0.7

Somatostatin-positive area (%) (n = 8, 3, 4, and 5, respectively) 0.5 6 0.1 0.6 6 0.2 0.2 6 0.1 0.7 6 0.3

PP-positive area (%) (n = 8, 3, 4, and 5, respectively) 0.6 6 0.1 0.1 6 0.1* 0.2 6 0.1 0.2 6 0.1

Data are mean6 SEM or n unless otherwise indicated. *P, 0.05 vs. CF-no DM, no transplant group.

care.diabetesjournals.org Hull and Associates 827

Unexpectedly, in contrast to some pre-vious studies, we did not observe a de-crease in relative b-cell area in subjectswith CFRD relative to subjects with CF

without diabetes (17–19) or control sub-jects without CF (20). However, the latterstudy used an older cohort of control sub-jects without CF (mean age 65 years) as a

comparison group. This could explain thediscrepant findings; specifically, oldersubjects are likely to have a higher BMIthan younger subjects (especially those

Figure 2—Quantitation of islet endocrine cell types in subjects with and subjects without CF. Insulin area (A) and glucagon area (B) normalized to totaltissue area. Linear regression analysis for the relationship between relative insulin and glucagon areas in non-CF (younger) control subjects (C), non-CF(older) control subjects (D), and CF-no DM (E) and CFRD (F) subjects. In a subset of subjects (n = 10 per group), somatostatin area (G) and pancreaticpolypeptide area (H) normalized to total tissue area.

828 Islet IL-1b Immunoreactivity in Cystic Fibrosis Diabetes Care Volume 41, April 2018

with CF), and increased BMI is associatedwith increased islet b-cell area (32). Forthis reason, we were careful to age, sex,and BMI match our comparison groupswithout CF and without diabetes to avoidthis potential confounder. Those previousstudies that reported decreased b-cellarea between subjects with CF, withand without diabetes, included smallnumbers of subjects with CFRD (n = 6–7)(17,18), whereas a larger study (n =16 subjects with CFRD) also failed tofind a significant difference in b-cellarea between subjects with CF with andsubjects with CF without diabetes (20). Itshould be noted that without pancreaticmass data (which has been shown to bedecreased in subjectswith CF [33]), it is notpossible to definitively determinewhetherb-cell mass is preserved in CF/CFRD.However, our present data, in agreementwith those of Couce et al. (20), suggestthat a predominantly functional defectunderlies the impaired insulin secretionseen in subjects with CFRD, which wouldbe supported by animal data showing thatCF pigs exhibit reduced insulin release de-spite preservation of islet mass (16).We observed a striking increase in islet

a-cell area in both CF-no DM and CFRDgroups compared with non-CF controlsubjects. The source of these new a-cellsis unknown but could arise because ofproliferation of existinga-cells, formationof new a-cells from ductal or other pre-cursors, or transdifferentiation of b-cells,the latter having been proposed as asource of increased a-cell mass in otherforms of diabetes or models thereof (34).These data are intriguing, given that sub-jects with CF exhibit impaired glucagonresponses to stimuli such as hypoglyce-mia (35), suggesting thata-cell expansionmay occur as an attempt to compensatefor decreaseda-cell function. Conversely,these morphometric data correlate withprevious clinical studies that have docu-mented impaired suppression of gluca-gon responses to oral glucose tolerancetests in CFRD subjects (13,36) and in vitrodata showing that cystic fibrosis trans-membrane conductance regulator (CFTR)is a negative regulator of glucagon release(37), situations where glucagon levelswould be increased under conditions ofCF. Interestingly, in subjects with CF with-out diabetes, glucagon area was corre-lated with insulin area, suggesting thatin the face ofa-cell expansion, islet endo-crine cell composition was still somewhat

preserved in this group. In contrast, therewas no such relationship between gluca-gon and insulin areas in subjects withCFRD, suggesting that abnormal isletcomposition characterizes diabetes in CFsubjects.

The finding of IL-1b in the islets of CFand CFRD subjects suggests an activationof the NLRP3 inflammasome in islet cells(likely b-cells). The NLRP3 inflammasomeis an intracellular protein complex thatresponds to a variety of intracellular andextracellular signals by triggering inflam-matory cascades, resulting in IL-1b ex-pression (38,39). While we did observeimmune cell infiltrates in exocrine pan-creas of subjects with CF, these were notpresent within islets aside from one veryyoung subject, suggesting that they werenot the source of the intra-islet IL-1b im-munoreactivity. Our current data do notreveal the mechanism of activation of in-flammasome activation in CF subjects.One hypothesis could be that mutationsin the CFTR could generate intracellularstress in b-cells, resulting in inflamma-some activation. Another possibility isthat extensive exocrine tissue autolysisin CF subjects could prime and initiateinflammasome-mediated IL-1b produc-tion. Finally, systemic inflammatory pro-cesses in subjects with CF appear skewedto a program of IL-1b–regulated cyto-kines (e.g., IL-8, IL-17 [26–28]), suggestingthat circulating levels of IL-1b itself mayalso be increased. This could in turn resultin autostimulation of IL-1b expression inislets, a phenomenon that has beendocu-mented in cultured human islets (40). Theislet in CF is therefore also likely exposedto increased IL-1b from both local andsystemic sources, making it especially vul-nerable to the effects of this proinflam-matory cytokine. Additional studies areneeded to investigate themechanisms un-derlying islet IL-1b production in CF/CFRD.

The toxic actions of IL-1b can be atten-uated by its endogenous antagonist,IL-1Ra. In the context of type 2 diabetes,it has been suggested that individualswithout diabetes express high levels ofIL-1Ra in islets, whereas subjects with di-abetes exhibit a loss of IL-1Ra, further ex-acerbating the proinflammatory effects ofIL-1b (30). In the current study, we wereable to detect only small numbers ofIL-1Ra–positive cells in islets from any sub-ject, regardless of CF or diabetes status.This excludes thepossibility that anyeffectsof islet IL-1b production (for example, in

CF subjects without diabetes) may be off-set by increased islet IL-1Ra and suggeststhe possibility of a targeted interventionwith clinically available IL-Ra antagonists.

Our study has some limitations. First,premortem oral glucose tolerance testdata were not available for the vast ma-jority of subjects, precluding the deter-mination of glucose tolerance status,particularly in the CF-no diabetes group.Further, our analysis of inflammatory cy-tokine immunoreactivity was limited toIL-1b, and our investigation into islet hor-mone immunoreactivity in somecaseswaslimited to 10 subjects per group. In addi-tion, given the limited number of subjects,we couldnot adequately control for poten-tially confounding factors such as concom-itant medication usage. However, despitethe presence of these potential confound-ers we were still able to identify islet IL-1bimmunoreactivity as a common featureof CF and exclude islet amyloid as a likelycause of its presence. Finally, the cross-sectional nature of the analysis of postmor-temspecimens limitsourabilitytodefinitivelycharacterize a sequential process in whichIL-1b production occurs before islet amy-loidosis. However, the development ofmore refined animal models and the useof IL-1–related biomarkers may be usefulin answering these questions both in thelaboratory and at the bedside.

In conclusion, we provide a histologicanalysis of human CF pancreas speci-mens, in subjects with or without diabe-tes, in which we identified islet IL-1bimmunoreactivity as well as increasedglucagon immunoreactivity as key abnor-malities in islet morphology. These find-ings may help explain islet failure andobserved metabolic abnormalities in pa-tients with CF. Future studies shouldfocus on a better understanding of thephysiologic effects of CFTR on b-cell bi-ology as well as a broader exploration ofinflammasome activation in the islet ofsubjects with CF. In addition, the develop-ment of IL-1b–related peripheral bio-markers may be an important tool tohelp in the early detection of CFRD. Thesefindings also suggest a possible role foranti-inflammatory drugs in preventingthe development of islet pathology in CF.

Acknowledgments. The authors thank SonyaHeltshe (Clinical Core, University of WashingtonCystic Fibrosis Research and Translation Center)for biostatistical consultation and Brian Johnson

care.diabetesjournals.org Hull and Associates 829

and Megan Larmore (Host Response Core, Uni-versity of Washington Cystic Fibrosis Researchand Translation Center) for expert assistancewith IHC and morphometric analyses.Funding. This research was performed with thesupport of nPOD, a collaborative type 1 diabetesresearch project sponsored by JDRF. Organ pro-curement organizations partnering with nPOD toprovide research resources are listed at www.jdrfnpod.org/for-partners/npod-partners/. Thiswork was supported by the University of Wash-ington Cystic Fibrosis Research and TranslationCenter (National Institute of Diabetes and Diges-tive and Kidney Diseases grant P30-DK-089507),through a Pilot and Feasibility Award (to R.L.H.and S.S.) and through work performed in theHost Response and Clinical Cores. This work wasalso supported by the University of Washington’sInstitute of Translational Health Sciences, whichreceives support from the National Center forAdvancing Translational Sciences (UL1 TR000423)and Cystic Fibrosis Foundation Research Develop-mentProgramawards (R565CR11, SINGH15R0, andVERKMA15R0).Duality of Interest. No potential conflicts of in-terest relevant to this article were reported.AuthorContributions.R.L.H.,R.L.G.,C.W.F.,B.W.R.,and S.S. conceived the study and designed theexperiments. R.L.H., R.L.G., S.M., G.H.D., C.L.F.,C.W.F., and S.S. collected and analyzed data. R.L.H.and S.S. performed data analysis and wrote themanuscript. R.L.G., G.H.D., C.L.F., C.W.F., and B.W.R.revised the manuscript. All authors approved thefinal version of the manuscript. R.L.H. and S.S. arethe guarantors of this work and, as such, had fullaccess to all the data in the study and take re-sponsibility for the integrity of the data and theaccuracy of the data analysis.Prior Presentation. Parts of this study were pre-sented in abstract format the 74th Scientific Sessionsof the American Diabetes Association, San Francisco,CA, 13–17 June 2014.

References1. MoranA,Dunitz J, NathanB, SaeedA,HolmeB,Thomas W. Cystic fibrosis-related diabetes: cur-rent trends in prevalence, incidence, and mortal-ity. Diabetes Care 2009;32:1626–16312. Ode KL, Frohnert B, Laguna T, et al. Oral glu-cose tolerance testing in childrenwith cystic fibro-sis. Pediatr Diabetes 2010;11:487–4923. Ode KL, Moran A. New insights into cysticfibrosis-related diabetes in children. Lancet Dia-betes Endocrinol 2013;1:52–584. Yi Y, Norris AW, Wang K, et al. Abnormal glu-cose tolerance in infants and young children withcystic fibrosis. Am J Respir Crit Care Med 2016;194:974–9805. Moran A, Brunzell C, Cohen RC, et al.; CFRDGuidelines Committee. Clinical care guidelines forcystic fibrosis-related diabetes: a position state-ment of the American Diabetes Association and aclinical practice guideline of the Cystic FibrosisFoundation, endorsed by the Pediatric EndocrineSociety. Diabetes Care 2010;33:2697–27086. Lanng S, Thorsteinsson B, Pociot F, et al. Diabe-tes mellitus in cystic fibrosis: genetic and immuno-logical markers. Acta Paediatr 1993;82:150–1547. Konrad K, Kapellen T, Lilienthal E, et al.; DPVInitiative and the Competence Network DiabetesMellitus. Does b-cell autoimmunity play a role

in cystic fibrosis-related diabetes? Analysis basedon the German/Austrian Diabetes PatientenVerlaufsdokumentation Registry. Diabetes Care2016;39:1338–13448. Blackman SM, Commander CW, Watson C,et al. Genetic modifiers of cystic fibrosis-relateddiabetes. Diabetes 2013;62:3627–36359. Blackman SM, Hsu S, Ritter SE, et al. A suscep-tibility gene for type 2 diabetes confers substan-tial risk for diabetes complicating cystic fibrosis.Diabetologia 2009;52:1858–186510. Furgeri DT, Marson FA, Ribeiro AF, BertuzzoCS. Associationbetween the IVS4G.Tmutation inthe TCF7L2 gene and susceptibility to diabetes incystic fibrosis patients. BMC Res Notes 2012;5:56111. Austin A, Kalhan SC, Orenstein D, Nixon P,Arslanian S. Roles of insulin resistance and beta-cell dysfunction in the pathogenesis of glucoseintolerance in cystic fibrosis. J Clin EndocrinolMetab 1994;79:80–8512. Elder DA, Wooldridge JL, Dolan LM, D’AlessioDA. Glucose tolerance, insulin secretion, and in-sulin sensitivity in children and adolescents withcystic fibrosis and no prior history of diabetes. JPediatr 2007;151:653–65813. Lanng S, Thorsteinsson B, Røder ME, Nerup J,Koch C. Insulin sensitivity and insulin clearance incystic fibrosis patients with normal and diabeticglucose tolerance. Clin Endocrinol (Oxf) 1994;41:217–22314. Olivier AK, Yi Y, Sun X, et al. Abnormal endo-crine pancreas function at birth in cystic fibrosisferrets. J Clin Invest 2012;122:3755–376815. Rogers CS, Stoltz DA, Meyerholz DK, et al.Disruption of the CFTR gene produces a modelof cystic fibrosis in newborn pigs. Science 2008;321:1837–184116. Uc A, Olivier AK, Griffin MA, et al. Glycaemicregulation and insulin secretion are abnormal incystic fibrosis pigs despite sparing of islet cellmass. Clin Sci (Lond) 2015;128:131–14217. Soejima K, Landing BH. Pancreatic islets inolder patients with cystic fibrosis with andwith-out diabetes mellitus: morphometric and im-munocytologic studies. Pediatr Pathol 1986;6:25–4618. Abdul-Karim FW, Dahms BB, Velasco ME,Rodman HM. Islets of Langerhans in adolescentsand adults with cystic fibrosis. A quantitativestudy. Arch Pathol Lab Med 1986;110:602–60619. Iannucci A, Mukai K, Johnson D, Burke B. En-docrine pancreas in cystic fibrosis: an immunohis-tochemical study. Hum Pathol 1984;15:278–28420. Couce M, O’Brien TD, Moran A, Roche PC,Butler PC. Diabetes mellitus in cystic fibrosis ischaracterized by islet amyloidosis. J Clin Endocri-nol Metab 1996;81:1267–127221. Clark A,Wells CA, Buley ID, et al. Islet amyloid,increased A-cells, reduced B-cells and exocrinefibrosis: quantitative changes in the pancreas intype 2 diabetes. Diabetes Res 1988;9:151–15922. Jurgens CA, Toukatly MN, Fligner CL, et al.b-Cell loss and b-cell apoptosis in human type 2diabetes are related to islet amyloid deposition.Am J Pathol 2011;178:2632–264023. Westermark P. Quantitative studies on amy-loid in the islets of Langerhans. Ups J Med Sci1972;77:91–9424. Masters SL, Dunne A, Subramanian SL, et al.Activation of the NLRP3 inflammasome by isletamyloid polypeptide provides a mechanism for

enhanced IL-1b in type 2 diabetes. Nat Immunol2010;11:897–90425. Westwell-Roper C, Dai DL, Soukhatcheva G,et al. IL-1 blockade attenuates islet amyloid poly-peptide-induced proinflammatory cytokine re-lease and pancreatic islet graft dysfunction. JImmunol 2011;187:2755–276526. Mulcahy EM, Hudson JB, Beggs SA, Reid DW,Roddam LF, Cooley MA. High peripheral bloodth17 percent associated with poor lung functionin cystic fibrosis. PLoS One 2015;10:e012091227. Tan HL, Regamey N, Brown S, Bush A, LloydCM, Davies JC. The Th17 pathway in cystic fibrosislung disease. Am J Respir Crit CareMed 2011;184:252–25828. Tiringer K, Treis A, Fucik P, et al. A Th17- andTh2-skewedcytokineprofile in cysticfibrosis lungsrepresents a potential risk factor for Pseudomo-nas aeruginosa infection. Am J Respir Crit CareMed 2013;187:621–62929. Rountree AM, Reed BJ, Cummings BP, et al.Loss of coupling between calcium influx, energyconsumption and insulin secretion associatedwith development of hyperglycaemia in theUCD-T2DM rat model of type 2 diabetes. Diabe-tologia 2013;56:803–81330. Maedler K, Sergeev P, Ehses JA, et al. Leptinmodulates beta cell expression of IL-1 receptorantagonist and release of IL-1beta in human islets.Proc Natl Acad Sci U S A 2004;101:8138–814331. CnopM,Welsh N, Jonas JC, Jorns A, Lenzen S,Eizirik DL. Mechanisms of pancreatic beta-celldeath in type 1 and type 2 diabetes: many differ-ences, few similarities. Diabetes 2005;54(Suppl. 2):S97–S10732. Butler AE, Janson J, Bonner-Weir S, Ritzel R,Rizza RA, Butler PC. Beta-cell deficit and increasedbeta-cell apoptosis in humans with type 2 diabe-tes. Diabetes 2003;52:102–11033. Sequeiros IM, Hester K, Callaway M, et al.;Bristol Cystic Fibrosis DiabetesGroup.MRI appear-ance of the pancreas in patients with cystic fibro-sis: a comparison of pancreas volume in diabeticand non-diabetic patients. Br J Radiol 2010;83:921–92634. Talchai C, Xuan S, Lin HV, Sussel L, Accili D.Pancreaticb cell dedifferentiation as amechanismof diabeticb cell failure. Cell 2012;150:1223–123435. Moran A, Diem P, Klein DJ, Levitt MD,Robertson RP. Pancreatic endocrine function incystic fibrosis. J Pediatr 1991;118:715–72336. Hinds A, Sheehan AG, Machida H, ParsonsHG. Postprandial hyperglycemia and pancreaticfunction in cystic fibrosis patients. Diabetes Res1991;18:69–7837. Edlund A, Pedersen MG, Lindqvist A, WierupN, Flodstrom-Tullberg M, Eliasson L. CFTR is in-volved in the regulation of glucagon secretion inhuman and rodent alpha cells. Sci Rep 2017;7:9038. Vanaja SK, RathinamVA, Fitzgerald KA.Mech-anisms of inflammasome activation: recent ad-vances and novel insights. Trends Cell Biol 2015;25:308–31539. Guo H, Callaway JB, Ting JP. Inflammasomes:mechanism of action, role in disease, and thera-peutics. Nat Med 2015;21:677–68740. Boni-SchnetzlerM, Thorne J, ParnaudG, et al.Increased interleukin (IL)-1beta messenger ribo-nucleic acid expression in beta -cells of individualswith type 2 diabetes and regulation of IL-1beta inhuman islets by glucose and autostimulation. JClin Endocrinol Metab 2008;93:4065–4074

830 Islet IL-1b Immunoreactivity in Cystic Fibrosis Diabetes Care Volume 41, April 2018