Page 1
1
1
ENHANCED GLUCOSE TOLERANCE IN PANCREATIC DERIVED 2
FACTOR (PANDER) KNOCKOUT C57BL/6 MICE 3
4
Shari L. Moak1, Grace C. Dougan2, Catherine B. MarElia1, Whitney A. Danse1, Amanda M. 5
Fernandez1, Melanie N. Kuehl1, Mark G. Athanason1 and Brant R. Burkhardt1* 6
7
1Department of Cell Biology, Microbiology and Molecular Biology, University of South Florida, 8
4202 East Fowler Avenue, Tampa, FL 33620. 9
2Department of Pediatrics, University of South Florida, 12901 Bruce B. Downs Blvd, MDC 62, 10
Tampa, FL 33612. 11
12
* Corresponding Author: Brant R. Burkhardt, Ph.D., Department of Cell Biology, Microbiology 13
and Molecular Biology, University of South Florida, 4202 East Fowler Avenue, BSF 14
206,Tampa, FL 33620, Tel: 813-974-5968, Fax: 813-974-1644, Email: [email protected] 15
16
SHORT TITLE: ENHANCED MODEL OF PANDER 17
18
Word count from abstract to discussion: 5131 (not including references) 19
20
6 Figures 21
22
23
© 2014. Published by The Company of Biologists Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed.
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
http://dmm.biologists.org/lookup/doi/10.1242/dmm.016402Access the most recent version at DMM Advance Online Articles. Posted 12 September 2014 as doi: 10.1242/dmm.016402
http://dmm.biologists.org/lookup/doi/10.1242/dmm.016402Access the most recent version at First posted online on 12 September 2014 as 10.1242/dmm.016402
Page 2
2
ABSTRACT 24
PANcreatic-DERived Factor (PANDER, FAM3B) is a uniquely structured protein strongly 25
expressed within and secreted from the endocrine pancreas. PANDER has been hypothesized to 26
regulate fasting and fed glucose homeostasis, hepatic lipogenesis and insulin signaling, and serve 27
a potential role in the onset or progression of type 2 diabetes. Despite having a potential 28
pleiotropic pivotal role in glycemic regulation and T2D, there has been limited generation of 29
stable animal models for PANDER investigation, with none on well-established genetic murine 30
backgrounds for T2D. Our aim was to generate an enhanced murine model to further elucidate 31
the biological function of PANDER. Therefore, a pure bred PANDER C57BL/6 knockout model 32
(PANKO-C57) was created and phenotypically characterized with respect to glycemic regulation 33
and hepatic insulin signaling. The PANKO-C57 exhibited an enhanced metabolic phenotype 34
particularly with regard to enhanced glucose tolerance. Male PANKO-C57 mice displayed 35
decreased fasting plasma insulin and c-peptide levels, whereas leptin levels were increased as 36
compared to matched C57BL/6J WT mice. Despite similar peripheral insulin sensitivity between 37
both groups, hepatic insulin signaling was significantly increased during fasting conditions as 38
demonstrated by increased phosphorylation of hepatic Akt and AMPK along with mature 39
SREBP-1 expression. Insulin stimulation of PANKO-C57 mice resulted in increased hepatic 40
triglyceride and glycogen content as compared to C57BL/6 WT. In summary, the PANKO-C57 41
mouse represents a suitable model for the investigation of PANDER in multiple metabolic states 42
and provides an additional tool to elucidate the biological function and potential role in T2D. 43
Abstract word count: 241 44
KEY WORDS: Pancreatic-derived factor, FAM3B, knockout model, glycemic regulation, 45
hepatic insulin signaling, glucose tolerance 46
47
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 3
3
INTRODUCTION 48
49
Pancreatic-Derived Factor (PANDER, FAM3B), discovered in 2002, is a 235-amino acid protein 50
that belongs to the family with sequence similarity 3 (FAM3) gene family (Zhu et al., 2002). 51
This cytokine-like gene family (based on predicted secondary structure) has four members 52
described as FAM3A, FAM3B, FAM3C, and FAM3D. FAM3B was later named PANDER 53
since this hormone is robustly expressed in and secreted from the endocrine pancreas (α and β 54
cells, specifically) and to a lesser extent other tissues such as the liver, small intestine and 55
prostate (Li et al., 2011; Li et al., 2013; Zhu et al., 2002). From the ostensible recognition of 56
folds (ORF) algorithm that was used in PANDER’s initial discovery, it was determined that 57
PANDER (then identified as FAM3B) had a predicted typical 4-helix bundle secondary structure 58
present in many known cytokines (Aurora and Rose, 1998; Zhu et al., 2002). However, more 59
recently elucidated crystal structures of murine secreted PANDER revealed a novel and unique 60
globular β-β-α fold (Johansson et al., 2013). PANDER is comprised of two anti-parallel beta 61
sheets lined by three short helices composed to form a highly conserved water-filled cavity that 62
does not resemble any currently known cytokine and thereby contradicting the earlier predictive 63
models (Johansson et al., 2013). This structure was conserved among other members of the 64
FAM3 family. The homologous FAM3C, also known as interleukin-like EMT inducer (ILEI), 65
ILEI that has been previously demonstrated to be involved with epithelial-mesenchymal 66
transition demonstrated a similar structure (Lahsnig et al., 2009; Song et al., 2013; Waerner et 67
al., 2006). Therefore, PANDER and this gene family may be comprised of a novel class of 68
signaling molecules acting distinctly different from that of other known cytokines or hormones. 69
Initial in-vitro studies regarding PANDER demonstrated a potential role in glycemic regulation 70
based on mode of regulation. Glucose has been shown to significantly enhance PANDER 71
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 4
4
promoter activity and secretion from β cells and pancreatic islets (Burkhardt et al., 2005; Wang 72
et al., 2008; Yang et al., 2005). In addition, PANDER is co-secreted with insulin in response to 73
glucose (Carnegie et al., 2010; Xu et al., 2005). 74
PANDER’s mechanism of action and full biological effect has yet to be fully elucidated in-vivo 75
and this is due to the lack of appropriate permanent and genomically integrated rodent models. 76
The initial PANDER knockout (PANKO) model was generated on a mixed genetic background 77
and demonstrated glucose intolerance in the presence of enhanced hepatic insulin sensitivity 78
(HIS) with no observed differences in peripheral insulin sensitivity or fasting glycemic levels 79
(Robert-Cooperman et al., 2010). Despite enhanced HIS as determined by hyperinsulinemic-80
euglycemic clamp (HEC) studies, no further characterization was performed to identify the 81
causative hepatic molecules or pathway for this finding and thereby leaving a gap in the 82
understanding of hepatic PANDER action during both fasting and fed conditions. The targeted 83
disruption of PANDER impaired pancreatic islet insulin secretion as demonstrated by an 84
inhibited glucose-induced response by PANKO islets identified during islet perifusion and 85
calcium imaging studies. In contrast, neither endocrine pancreatic morphology nor insulin 86
content was affected by absence of PANDER. The lack of further hepatic molecular 87
characterization and a robust phenotype from the mixed genetic background PANKO mouse led 88
us to speculate that the lack of a congenic background via a pure-bred mouse model may be 89
hindering the penetrance of the phenotype and confounding the results particularly in 90
relationship to hepatic insulin sensitivity. Furthermore, a series of review articles have strongly 91
suggested that PANDER may potentially be implicated in the onset or progression of type 2 92
diabetes (T2D) (Wang et al., 2012; Wilson et al., 2011; Yang and Guan, 2013), yet the impact of 93
this gene within well-established models of T2D or insulin resistance has not been generated. 94
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 5
5
Therefore, to further refine and define the biological function of PANDER within the context of 95
a well-established model of T2D within a permissive genetic background, we generated and 96
characterized the PANDER knockout mouse on a pure C57BL/6 genetic background. 97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 6
6
RESULTS 138
139
Decreased fasting glycemia in PANKO-C57 mice 140
Prior to experimentation, all mice were genotyped as a confirmatory measure following receipt 141
of PANKO-C57 mice from Jackson Laboratories using primers described in Materials and 142
Methods (data not shown). The effect of PANDER on fasting glycemic levels was determined by 143
measuring blood glucose following a short (4 hours) and long-term (16 hours) fast. PANKO-144
C57 male mice aged 4 months displayed significantly decreased blood glucose levels after a 145
long-term fast compared to age and sex matched C57BL/6 mice (P < 0.01) (Fig. 1A). Decreased 146
fasting glycemia was also observed following a short-term fast with blood glucose values of 147
169.3 mg/dL + 6.9 versus 225.8 mg/dL + 30.5 (P < 0.05) (Fig. 1B). A similar but non-statistical 148
trend was observed with female PANKO-C57 long and short-term fasting mice (Figs. 1C and 149
1D). In summary, the absence of PANDER promoted decreased fasting glycemic levels that was 150
not previously observed or reported in the mixed genetic PANDER knockout or acutely 151
delivered PANDER siRNA models revealing a discernible phenotype within this background (Li 152
et al., 2011; Robert-Cooperman et al., 2010). 153
154
Enhanced glucose tolerance in PANKO-C57 155
To evaluate the effect of PANDER during postprandial conditions in this model, PANKO-C57 156
and WT mice of both genders were examined by glucose tolerance testing (GTT). GTTs 157
performed at 4 months of age revealed enhanced glucose tolerance in male PANKO-C57 mice 158
based on decreased glycemic levels throughout the duration of the GTT (Fig. 2A). Significantly 159
lower blood glucose measurements were observed in male PANKO-C57 compared to WT mice 160
during the course of the GTT (two-way ANOVA, P<0.05). Insulin tolerance tests (ITTs) were 161
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 7
7
performed to examine peripheral insulin sensitivity. Response to i.p. injected insulin was shown 162
to be similar between PANKO-C57 and WT mice (Fig. 2B). With regard to female PANKO-C57 163
mice, a similar trend of enhanced glucose tolerance was measured during the GTT but without 164
statistical significance (Fig. 2C). Insulin sensitivity during the ITT was similar between female 165
PANKO-C57 and WT mice (Fig. 2D). In summary, absence of PANDER promoted enhanced 166
glucose tolerance without significantly affecting peripheral insulin sensitivity in a male dominant 167
fashion. This was not observed in our prior PANKO mixed genetic background model and 168
demonstrates an enhanced metabolic phenotype in the pure C57BL/6 model (Robert-Cooperman 169
et al., 2010). 170
171
PANKO-C57 male and female mice have increased body weight 172
To further characterize the PANKO-C57, longitudinal body weight was evaluated from 8-23 173
weeks of age. Beginning at 8 weeks of age and fed normal chow ad-libitum, PANKO-C57 male 174
mice present with significantly increased weights compared to age and gender-matched WT mice 175
(up to 23 weeks of age (Fig. 3A). Measurements in body weight were terminated at 23 weeks of 176
age. This same trend was observed in female mice, and although significance diminishes over 177
time, PANKO-C57 female mice remain significantly heavier than WT mice at 21 weeks old 178
(Fig. 3B). Increased body weight has not been reported or observed in prior animal models and 179
reveals a potential prior overlooked function of PANDER in the realm of whole body 180
homeostasis. Furthermore, the appearance of this phenotype in both genders supports the 181
differentiation of this model from other PANDER animal surrogates (Li et al., 2011; Robert-182
Cooperman et al., 2010; Wilson et al., 2010). 183
184
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 8
8
Fasting hormonal levels in PANKO-C57 mice 185
Following the combined gender phenotyping, we focused our study on the male PANKO-C57 to 186
identify if hormonal differences accounted for the observed decreased fasting glycemic levels in 187
the male PANKO-C57. Therefore, fasting plasma levels of insulin, c-peptide, glucagon, leptin, 188
and corticosterone were measured at 2 and 5 months of age in male mice (Fig. 4). At 2 months of 189
age, PANKO-C57 male mice presented with significantly reduced insulin levels (Fig. 4A). 190
Glucagon levels were similar between PANKO-C57 and WT 2 and 5 month old mice (Fig. 4B). 191
Interestingly, leptin was increased in PANKO-C57 mice in both age groups (Fig. 4C). C-peptide 192
levels at 2 months of age were significantly decreased and concordant with observed decreased 193
measured insulin levels (Fig. 4D). Fasting corticosterone levels were similar between PANKO-194
C57 and WT mice in both age groups (Fig. 4E). In general, the hormonal results from fasting 195
plasma revealed characteristics of the PANKO-C57 model not observed in prior investigations 196
such as with regard to decreased insulin levels in 2 month old mice and increased leptin levels 197
(Li et al., 2011; Robert-Cooperman et al., 2010). In addition, glucose stimulated insulin levels 198
were decreased in the PANKO-57 with statistical significance at the conclusion of the glucose 199
tolerance test (Figure 4F). This result is consistent with prior reports indicating that the absence 200
of PANDER can impair glucose stimulated insulin secretion (Robert-Cooperman et al., 2010; 201
Robert-Cooperman et al., 2011). 202
Enhanced insulin-stimulated hepatic glycogen and triglyceride content 203
To elucidate further downstream metabolic consequences of PANDER, we examined hepatic 204
glycogen and triglyceride levels following insulin-stimulated and fasting conditions in male 205
mice. In terms of hepatic glycogen content, 15 minutes after stimulation with 2 units of insulin 206
per kg of body weight, the glycogen concentration within PANKO-C57 mouse livers was 207
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 9
9
significantly increased, over 120 fold, compared to that of WT mice (Fig 5A). Hepatic glycogen 208
levels were not significantly different following fasting conditions in the PANKO-C57 model as 209
compared to WT controls (Fig. 5A, right panel). Hepatic triglyceride concentration was also 210
increased in PANKO-C57 mice during insulin stimulated conditions as compared to WT mice 211
(Fig. 5B). During fasting conditions, hepatic triglyceride concentration was significantly 212
decreased in PANKO-C57 mice as compared to WT counterparts (Fig. 5B). Circulating plasma 213
triglyceride concentration was measured after an overnight fast and was significantly increased 214
in 2 month old PANKO mice as compared to WT mice but not at 5 months of age (Fig. 5C). In 215
summary, the liver responds very well to insulin in the PANKO-C57 model when it is stimulated 216
for the production of glycogen but not for triglyceride production. In addition, the lowered 217
fasting glucose levels suggest that hepatic gluconeogenesis is well suppressed by insulin. 218
219
PANKO-C57 male mice display enhanced hepatic signaling 220
Prior investigations by our laboratory and others have provided evidence that the liver is a major 221
target tissue of PANDER (Li et al., 2011; Robert-Cooperman et al., 2010; Robert-Cooperman et 222
al., 2014; Wilson et al., 2010). In addition, HEC studies performed in the former PANKO mouse 223
model on mixed genetic background indicated enhanced hepatic insulin sensitivity but the 224
mechanism and critical signaling molecules accounting for this were not characterized or 225
identified (Robert-Cooperman et al., 2010). Therefore, we evaluated the effect of PANDER on 226
critical hepatic signaling molecules involved in canonical insulin signaling and triglyceride 227
synthesis within the PANKO-C57 model. Western blot analysis of numerous hepatic insulin 228
signaling molecules were performed on male PANKO-C57 and WT livers excised after an 229
overnight fast or insulin stimulation (Fig. 6). 230
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 10
10
During fasting conditions, increased levels of phosphorylated hepatic signaling molecules were 231
observed in the PANKO-C57 model (Fig. 6). Significantly increased phosphorylated levels of 232
PI-3K, Akt and adenosine monophosphate-activated protein kinase (AMPK) were observed 233
during fasting conditions (Fig. 6A-C). Furthermore, significant increased levels of mature 234
SREBP-1 were also observed and may potentially account for a mechanistic explanation for 235
increased triglyceride production (Fig. 6D). During insulin stimulatory conditions, increased 236
levels phosphorylated PI-3K, Akt, and AMPK were observed in the livers of PANKO-C57 as 237
compared WT controls (data not shown). However, following densitometric analysis, differences 238
were not statistically significant (data not shown). Taken together, hepatic signaling in the 239
PANKO-C57 model demonstrated an overall enhancement of critical insulin signaling 240
particularly in the fasted state which is a unique observation within the PANKO-C57 model and 241
suggests PANDER serving a potential role in regulation of hepatic signaling during both feeding 242
and fasting conditions. 243
244
245
246
247
248
249
250
251
252
253
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 11
11
4. DISCUSSION 254
Distinct metabolic differences observed in the PANKO-C57 model as compared to mixed 255
genetic background 256
257
In contrast to the previous mixed PANKO model, the PANKO-C57 demonstrated: (1) enhanced 258
glucose tolerance during GTT and fasting conditions, (2) decreased fasting insulin and c-peptide 259
levels, (3) increased fasting leptin levels, (4) increased body weight and (5) increased p-Akt, p-260
AMPK and mature SREBP-1 levels. Our original characterization of the PANDER knockout was 261
performed on a mixed genetic background of C57BL/6 and 129Sv. The previous PANKO model 262
revealed the following differences as compared to WT controls: (1) glucose intolerance during 263
GTT but not fasting conditions, (2) both impaired insulin secretion and clearance, (3) no 264
differences in fasting glucagon, insulin, or leptin, (4) similar peripheral insulin sensitivity, (5) 265
enhanced hepatic insulin sensitivity, and (6) no differences in body weight (Robert-Cooperman 266
et al., 2010). Compared to the previous PANKO model, the PANKO-C57 displays a much more 267
exaggerated phenotype, particularly during fasting conditions, in terms of additional differences 268
from WT controls. Also, the PANTG and mixed PANKO models revealed distinct differences 269
within males only, whereas, the PANKO-C57 exhibited increased weight gain and matched 270
observed metabolic differences primarily seen in females, providing further evidence of a more 271
pronounced penetrance of phenotype on the C57BL/6 background. The influence of genetic 272
strain on phenotype has been well documented in mice (Berglund et al., 2008; Goren et al., 273
2004). This difference may be attributed to the affect that the host genome can have on 274
numerous metabolic factors such as with regard to insulin secretion or peripheral and hepatic 275
sensitivity (Doetschman, 2009; Kahle et al., 2013). This influence of genetic background has 276
been well documented particularly in the C57BL/6J mouse model, which serves as a common in-277
bred strain of diet-induced T2D (Freeman et al., 2006; Toye et al., 2005). The findings from the 278
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 12
12
PANKO-C57 model were both consistent and contrasting from the mixed genetic background. 279
The primary consistent phenotype from both models was with regard to hepatic insulin signaling. 280
Although increased hepatic insulin sensitivity was demonstrated in the mixed PANKO model, 281
there was no reported characterization of critical signaling molecules such as Akt or AMPK. 282
However and consistent with the observed hepatic insulin sensitivity reported in the prior model, 283
the PANKO-C57 demonstrated an enhancement of both hepatic Akt and AMPK phosphorylation 284
in addition to other downstream effectors such as SREBP-1. Furthermore, another concordant 285
observation was with peripheral insulin sensitivity being unaffected in both knockout models. 286
Taken together, the overall hypothesis that PANDER impacts glycemic levels in part through 287
interaction with the liver and hepatic signaling was certainly supported. Also, the PANTG and 288
mixed PANKO models revealed distinct differences within males only, whereas, the PANKO-289
C57 exhibited increased weight gain and matched observed metabolic differences in both males 290
and females, providing further evidence of a more pronounced penetrance of phenotype on the 291
C57BL/6J background. The prior PANKO mixed model did report impaired insulin secretion 292
resulting in glucose intolerance. This matched observation was also supported somewhat in the 293
PANKO-C57 model as decreased fasting levels of insulin and c-peptide were reported but this 294
did not have any impact on glucose intolerance as previously reported in the mixed model. It is 295
reasonable to speculate that perhaps the impaired insulin secretion is adequately compensated by 296
enhanced hepatic insulin sensitivity and therefore does not result in an observable impaired 297
glucose response. 298
299
PANKO-C57 model vs. PANDER transgenic/acute models 300
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 13
13
During the preparation of this manuscript, we recently published the findings from our PANDER 301
transgenic model (PANTG) (Robert-Cooperman et al., 2014). This transgenic mouse exclusively 302
overexpresses PANDER from the pancreatic β-cell which results in an approximate 4-fold 303
increase in circulating PANDER levels in both the fasted and fed state. The phenotype of the 304
PANTG as compared to PANKO-C57 was strikingly concordant in terms of the following 305
observations: (1) fasting glucose, C-peptide, and insulin levels were elevated (decreased in 306
PANKO-C57), (2) hepatic insulin sensitivity was decreased, (3) hepatic p-AMPK was decreased 307
during insulin stimulatory conditions, and (4) glucose intolerance was observed during GTT 308
(Robert-Cooperman et al., 2014). In general, the consistency between both the observed 309
metabolic phenotype between the PANKO-C57 and PANTG strongly implicates a pleiotropic 310
role for PANDER in the disruption of hepatic insulin signaling and impacting glycemic levels in 311
both the fasted and fed state. Acute models of PANDER employing transient overexpression 312
within the liver via adenoviral delivery (Ad-PANDER) have also demonstrated concordant data 313
(Li et al., 2011; Wilson et al., 2010). Li et al. reported that overexpression of hepatic PANDER 314
resulted in suppressed Akt activation and hepatic silencing of PANDER decreased insulin 315
resistance in db/db mice, however did not reveal any significant differences in glycemic levels 316
during fasting or fed conditions (Li et al., 2011). Wilson et al. revealed that Ad-PANDER mice 317
demonstrated increased fasting glucose and insulin levels along with glucose intolerance (Wilson 318
et al., 2010). However, Wilson et al. concluded that the glucose intolerance was due to elevated 319
basal fasting glycemic levels rather than impaired insulin secretion or sensitivity. This effect was 320
attributed to PANDER amplifying glucagon signaling via hepatic cAMP and cAMP-response 321
element-binding protein in contrast to impaired Akt signaling reported by Li et al. The 322
observable and discernible phenotype within the PANKO-C57 with regard to both fasting and 323
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 14
14
fed glucose intolerance not seen in prior models provides a measurable outcome and suitable 324
metabolic endpoint to further investigate PANDER and enable defined outcomes with this model 325
not observed in prior studies. 326
327
Absence of PANDER resulting in increased leptin and hepatic insulin sensitivity 328
Another interesting finding observed uniquely in the PANKO-C57 model not found in prior 329
PANDER models is the increased serum leptin levels. Leptin is produced from adipocytes and 330
serves a critical role in both weight gain and appetite (Ahima and Flier, 2000). Gross differences 331
in food consumption were not observed in the PANKO-C57 model (SLM, personal observation). 332
Administration of leptin to ob/ob or C57BL/6J mice demonstrated improved hepatic insulin 333
sensitivity and decreased fasting glycemia and body weight (Halaas et al., 1995; Pelleymounter 334
et al., 1995). Circulating leptin levels are associated with increased body weight in both rodents 335
and humans (Considine et al., 1996; Frederich et al., 1995). In addition, leptin has been 336
demonstrated to enhance the suppression of critical gluconeogenic enzymes such as PEPCK. The 337
PANKO-C57 model was similar in most characteristics with leptin models in terms of decreased 338
fasting glycemia and enhanced hepatic signaling. The PANKO-C57 model demonstrated 339
increased leptin and weight gain in the presence of decreased fasting and glucose stimulated 340
insulin levels. This result may be explained in part by leptin promoting hepatic insulin sensitivity 341
and reducing the demand on insulin secretion by the pancreatic β-cell. In addition, PANDER has 342
been shown to facilitate insulin secretion and impair hepatic insulin sensitivity. The synergistic 343
impact of these effects could result in a pronounced insulin sensitive state in the presence of 344
impaired pancreatic β-cell function reducing overall circulating insulin levels. Further 345
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 15
15
experimentation is needed to determine if PANDER has a potential suppressive effect on leptin 346
production. 347
348
The potential impact of PANDER on type 2 diabetes 349
With accumulating investigations and review articles suggesting PANDER may influence T2D 350
and nonalcoholic fatty liver disease by manner of selective hepatic insulin resistance (Wang et 351
al., 2012; Wilson et al., 2011; Yang and Guan, 2013). The majority of previous reports from both 352
the stable and transient PANDER models suggests that PANDER disrupts insulin signaling, 353
particularly with Akt, yet hepatic lipogenesis is increased (Li et al., 2011; Robert-Cooperman et 354
al., 2014). Strong evidence has demonstrated that PI3K/Akt activates the sterol-regulatory 355
element-binding proteins (SREBPs), which are considered master transcriptional regulators of 356
lipid metabolism (Krycer et al., 2010). Therefore, PANDER induced disruption of hepatic 357
signaling still does not alter hepatic triglyceride production and demonstrates that PANDER can 358
result in a selective hepatic insulin resistant state whereby suppressed Akt signaling does not 359
impair lipogenesis but yet still fails to inhibit gluconeogenesis. Selective hepatic insulin 360
resistance is a major pathological aspect of T2D and responsible for both hyperglycemia and 361
dyslipidemia (Biddinger et al., 2008; Brown and Goldstein, 2008). The hepatic cellular 362
mechanism by which PANDER induces SHIR is largely unknown but has been suggested to be 363
either mediated potentially through inhibited phosphorylation of AMPK or via Forkhead box 1 364
(FOXO1). The findings in the PANKO-C57 certainly support the impact on hepatic signaling 365
and have revealed not only increased p-Akt levels but downstream molecules such as mature 366
SREBP-1 with concordant increases in hepatic triglyceride content. Therefore, both the PANKO 367
and PANTG models are revealing a similar phenotype but this may be reconciled due to the 368
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 16
16
speculation that absence of PANDER promotes an overall insulin sensitive state that stimulates 369
both glycogen and triglyceride production. Indeed, hyperinsulinemic-euglycemic clamp studies 370
in the prior PANKO model did reveal decreased gluconeogenic output (Robert-Cooperman et al., 371
2010). HEC performed on the PANKO-C57 model would have been highly useful but were not 372
conducted due to current technical limitations to perform these experiments. 373
374
Utility of the PANKO-C57 in the investigation of PANDER 375
Overall, the PANKO-C57 mouse provides an excellent model for the investigation of PANDER 376
and has revealed additional and novel findings further elucidating the metabolic function of 377
PANDER and role of this uniquely structured molecule in glycemic regulation and hepatic 378
insulin signaling. Identification of the PANDER receptor would be highly valuable in the 379
understanding of PANDER biology along with measurement of circulating PANDER during 380
various metabolic and pathophysiological conditions particularly those found with T2D. This 381
conducted study evaluating the PANKO-C57 model in combination with the investigations of 382
others is strongly alluding to PANDER being a potential therapeutic candidate for the treatment 383
of hepatic insulin resistance and steatosis typically associated with T2D. 384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 17
17
MATERIALS AND METHODS 399
400
Generation of PANDER Knockout C57BL/6 Mice (PANKO-C57) 401
The initial description of the targeted disruption of the PANDER (FAM3B) gene and generation 402
of the knockout was previously described and performed at the Children’s Hospital of 403
Philadelphia Research Institute (CHOP) by replacing the transcriptional start site and first 2 404
exons of the PANDER gene with neomycin (Robert-Cooperman et al., 2010). The PANDER 405
knockout mouse was then backcrossed with C57BL/6 mice for 7 generations at CHOP followed 406
by subsequent additional crossings to C57BL/6J mice following commercial donation as 407
described next. This strain (PANKO-C57) was subsequently donated to Jackson Laboratories 408
(Bar Harbor, ME) and is commercially available under the strain name B6.129S6-Fam3btm1Bkht/J 409
(Stock Number 013788). Homozygote PANKO-C57 breeding pairs were obtained from Jackson 410
Laboratory and shipped to the Moffitt Cancer Center’s Stabile Vivarium (Tampa, FL) for murine 411
colony generation. All mice were fed Purina normal chow and water ad libitum. All PANDER 412
knockout and C57BL/6J WT offspring were screened for the presence of the knockout or WT 413
genes, 500 and 800 bp, respectively, via PCR amplification of genomic DNA isolated from tail 414
tissue (DNeasy Kit, Qiagen, Germantown, MD). PCR genotype confirmation cycling conditions 415
and primers (Forward primer located in PANDER promoter region: 5’- CTT GTG ATG GTG 416
GAT GCC CAG TT -3’ and Reverse primer located within neomycin gene: 5’- CTT CCT CGT 417
GCT TTA CGG TAT C -3’) were followed as previously described (Robert-Cooperman et al., 418
2010). Confirmed PANDER knockout and C57BL/6J WT mice aged 8 weeks to 6 months of age 419
were primarily evaluated for this study. All murine handling and experimentation adhered to 420
protocols approved by the Institutional Animal Care and Use Committee at the University of 421
South Florida and Moffitt Cancer Center. 422
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 18
18
Weight Measurements 423
PANKO-C57 and WT mice were weighed (Beckman Digital Scale) at the same time in the 424
morning (approximately 9 AM) during ad-libitum chow conditions in 2-3 week increments 425
beginning at 8 weeks of age for both genders. Weight was measured in grams and evaluated 426
longitudinally between both groups. 427
428
Glucose Tolerance Testing and insulin measurments 429
Glucose tolerance tests (GTTs) were performed on mice at 16 weeks of age as previously 430
described (Robert-Cooperman et al., 2010). In short, PANKO-C57 and WT mice were fasted 431
overnight (> 16 hrs) in a cage with bedding removed approximately 16 hours prior to the GTT. 432
Mice were then injected intraperitoneally with 2 grams of glucose (Fisher Scientific) per 433
kilogram of body weight. Blood glucose was measured at 0 minutes prior to the glucose injection 434
and at 15, 30, 60, and 120 minutes thereafter using a TRUEtrack® glucometer (Nipro 435
Diagnostics, Inc., Fort Lauderdale, FL) via a tail vein blood collection. Insulin levels were 436
measured using a commercially available murine insulin ELISA kit (ALPCO Diagnostics, 437
Salem, NH) during the glucose challenge. 438
439
Insulin Tolerance Testing 440
Insulin tolerance tests (ITTs) were performed within a week after GTT on mice at 16 weeks of 441
age as previously described (Robert-Cooperman et al., 2010). In brief, PANKO-C57 and WT 442
mice were fasted for 4 hours prior to the ITT. Mice were subsequently injected with NovoLog® 443
insulin (Novo Nordisk Inc., Plainsboro, NJ) at a concentration of 1U/kg of body weight. Blood 444
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 19
19
glucose was measured immediately before injection of insulin (0 minutes) and thereafter at 15, 445
30, 60, 90, and 120 minutes as described above for GTT. 446
447
Measurement of Fasting Glycemia 448
Mice were fasted either long term (overnight, approximately 16 hours) or short term (4 hours) 449
prior to metabolic testing as described above. Blood glucose was measured at the same time for 450
all mice from a tail vein collection as previously described. Blood glucose was averaged for each 451
group and compared based on fasting duration (n= 8-15). 452
453
Immunoblotting 454
Western blot analysis was performed on flash-frozen livers isolated from PANKO-C57 and WT 455
mice at 5-6 months of age. Mice were either fasted overnight (approximately 16 hours) or short-456
term (4 hours). For insulin stimulation experiments, mice were fasted short-term and then 457
intraperitoneally injected with insulin (NovoLog®, Novo Nordisk, Plainsboro, NJ) at 2 U/kg of 458
body weight. Protein was isolated from hepatic tissue using Tissue Protein Extracting Reagent 459
(TPER) (Thermo Fisher Scientific, Rockford, IL) and quantified using Pierce BCA Protein 460
Assay following manufacturer’s protocol (Thermo Fisher Scientific). Hepatic lysate (20-50 μg) 461
was analyzed using SDS-PAGE (Pre-Cast 10% Mini-PROTEAN® TGX™ gels, Bio-Rad, 462
Hercules, CA) and electrotransferred to polyvinylidine fluoride (PVDF) membrane using iBlot 463
semi-dry transfer apparatus (Invitrogen, Carlsbad, CA). Western blots were then probed with 464
antibodies for levels phosphorylated Phosphatidylinositol 3-Kinase (p-PI3K) at Tyr 508 (Cat. 465
Number sc-12929, Santa Cruz Biotechnologies, Inc., Santa Cruz, CA), total and phosphorylated 466
(position Thr 308) Akt/Protein Kinase B (PKB) (Cat. Number 4691 and 4056, Cell Signaling, 467
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 20
20
Danvers, MA), full-length mature and precursor Sterol Regulatory Element-Binding Protein 1 468
(SREBP-1) (Cat. Number sc-8984, Santa Cruz Biotechnologies, Inc.) and total and 469
phosphorylated (position Thr 172) Adenosine Monophosphate-Activated Protein Kinase subunit 470
α (AMPKα) (Cat. Number 2603 and 2535, Cell Signaling). All primary antibodies were diluted 471
1:1000 in StartBlock™ Block Buffer (Thermo Fisher Scientific). Protein detection was achieved 472
using horseradish peroxidase-conjugated goat-anti-rabbit secondary antibody (Bio-Rad) at 473
1:3000 dilution in StartBlock™ followed by chemiluminescence detection using Pierce ECL 474
Western Blotting Substrate (Thermo Scientific). Protein signals were visualized using the LAS 475
3000 Intelligent Dark Box (Fujifilm, Stamford, CT) and relative protein levels were normalized 476
to Glyceraldehyde 3-Phosphate Dehydrogenase (GAPDH) loading control (Cat. Number 2118, 477
Cell Signaling). Protein expression levels were quantified and compared using ImageJ (version 478
1.46) densitometry analysis. 479
480
Plasma Collection for Corticosterone/Triglyceride Assays and Endocrine Hormone Panel 481
482
PANKO-C57 and WT mice were fasting overnight prior to plasma collection. Fasting whole 483
blood was obtained via submandibular vein puncture and free flow collection into BD 484
MicroContainer® (Becton, Dickinson and Company, Franklin Lakes, NJ). Whole blood samples 485
were immediately placed on ice before being centrifuged for 5 minutes at 13.2 rpm for plasma 486
separation into a new Eppendorf (Eppendorf, Hauppague, NY) collection tube. The endocrine 487
hormone panel was performed using the commercially available Milliplex Mouse Metabolic 488
Hormone Magnetic Bead Panel MMHMAG-44K (EMD Millipore Corporation, Billerica, MA) 489
and Luminex Multiplex Platform (MAGPIX®). All procedures regarding the metabolic hormone 490
panel adhered to manufacturer’s protocol included in the assay kit. Corticosterone EIA kit (Enzo 491
Life Sciences, Farmingdale, NY) and Triglyceride Quantification Kit (Abcam®, Cambridge, 492
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 21
21
MA) were utilized for determination of corticosterone and plasma triglyceride concentration, 493
respectively, and both were performed in accordance with manufacturer’s suggested protocols. 494
495
Hepatic Glycogen Assay 496
PANKO-C57 and WT Mice were either fasted (overnight) or stimulated with insulin 497
(Novolog®) at a concentration of 2 U/kg of body weight prior to liver isolation. Livers were 498
harvested and immediately flash frozen in liquid nitrogen for storage at -80°C. 10 mg of liver 499
tissue was used with the Glycogen Assay Kit (ab65620) (Abcam®) according to manufacturer’s 500
suggested protocol. 501
502
Hepatic Triglyceride Assay 503
Overnight fasting and insulin-stimulated hepatic triglyceride content was evaluated in male 504
PANKO and WT mice at 4-5 months of age. Fasting and insulin-stimulated liver tissue harvest 505
was conducted as described above and the Triglyceride Quantification Kit (Abcam®) 506
colorimetric assay was conducted according to manufacturer’s protocol. 507
508
Statistical Analysis 509
Data are presented as mean + S.E.M. unless otherwise specified. All statistical analyses were 510
completed using GraphPad Prism (version 5) where statistical significance between groups was 511
determined by unpaired Student t-test or two-way ANOVA. Any P value that was less than 0.05 512
was considered to be statistically significant. 513
514
515
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 22
22
Acknowledgements 516
The authors also thank the Moffitt Stabile Vivarium for the continual maintenance of our animal 517
colony. The authors also acknowledge the tremendous contributions of the Jackson Laboratory 518
and in particular Melissa Osbourne and Jonathan Newell for providing critical mice for the 519
measurement of glucose stimulated insulin levels. The work presented here was in partial 520
fulfillment of the Master of Science degree in the department of Cell Biology, Microbiology and 521
Molecular Biology at the University of South Florida to S.L.M. 522
523
Competing Interests 524
The authors declare no competing financial interests. 525
526
Author contributions 527
S.L.M. designed and performed the experiments, analyzed the data and contributed to manuscript 528
writing. G.C.D. performed the experiments involving the metabolic evaluation of mice. K.B.M., 529
A.M.F., W.A.D., M.N.K., and M.G.A performed the experiments and greatly contributed to the 530
colony maintenance required for later phenotyping. B.R.B. designed the project, analyzed data, 531
and wrote the manuscript. 532
533
Funding 534
This work was partially supported by the USF New Investigator Award to BRB. 535
536
537
538
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 23
23
REFERENCES 539
540
Ahima, R. S. and Flier, J. S. (2000). Leptin. Annual Review of Physiology 62, 413-437. 541
Aurora, R. and Rose, G. D. (1998). Seeking an ancient enzyme in Methanococcus 542
jannaschii using ORF, a program based on predicted secondary structure comparisons. Proc Natl 543
Acad Sci U S A 95, 2818-23. 544
Berglund, E. D., Li, C. Y., Poffenberger, G., Ayala, J. E., Fueger, P. T., Willis, S. E., 545
Jewell, M. M., Powers, A. C. and Wasserman, D. H. (2008). Glucose metabolism in vivo in 546
four commonly used inbred mouse strains. Diabetes 57, 1790-1799. 547
Biddinger, S. B., Hernandez-Ono, A., Rask-Madsen, C., Haas, J. T., Aleman, J. O., 548
Suzuki, R., Scapa, E. F., Agarwal, C., Carey, M. C., Stephanopoulos, G. et al. (2008). 549
Hepatic insulin resistance is sufficient to produce dyslipidemia and susceptibility to 550
atherosclerosis. Cell Metab 7, 125-34. 551
Brown, M. S. and Goldstein, J. L. (2008). Selective versus total insulin resistance: a 552
pathogenic paradox. Cell Metab 7, 95-6. 553
Burkhardt, B. R., Yang, M. C., Robert, C. E., Greene, S. R., McFadden, K. K., 554
Yang, J., Wu, J., Gao, Z. and Wolf, B. A. (2005). Tissue-specific and glucose-responsive 555
expression of the pancreatic derived factor (PANDER) promoter. Biochim Biophys Acta 1730, 556
215-25. 557
Carnegie, J. R., Robert-Cooperman, C. E., Wu, J., Young, R. A., Wolf, B. A. and 558
Burkhardt, B. R. (2010). Characterization of the expression, localization, and secretion of 559
PANDER in alpha-cells. Mol Cell Endocrinol 325, 36-45. 560
Considine, R. V., Sinha, M. K., Heiman, M. L., Kriauciunas, A., Stephens, T. W., 561
Nyce, M. R., Ohannesian, J. P., Marco, C. C., McKee, L. J., Bauer, T. L. et al. (1996). 562
Serum immunoreactive-leptin concentrations in normal-weight and obese humans. N Engl J Med 563
334, 292-5. 564
Doetschman, T. (2009). Influence of genetic background on genetically engineered 565
mouse phenotypes. Methods Mol Biol 530, 423-33. 566
Frederich, R. C., Lollmann, B., Hamann, A., Napolitano-Rosen, A., Kahn, B. B., 567
Lowell, B. B. and Flier, J. S. (1995). Expression of ob mRNA and its encoded protein in 568
rodents. Impact of nutrition and obesity. J Clin Invest 96, 1658-63. 569
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 24
24
Freeman, H. C., Hugill, A., Dear, N. T., Ashcroft, F. M. and Cox, R. D. (2006). 570
Deletion of nicotinamide nucleotide transhydrogenase - A new quantitive trait locus accounting 571
for glucose intolerance in C57BL/6J mice. Diabetes 55, 2153-2156. 572
Goren, H. J., Kulkarni, R. N. and Kahn, C. R. (2004). Glucose homeostasis and tissue 573
transcript content of insulin signaling intermediates in four inbred strains of mice: C57BL/6, 574
C57BLKS/6, DBA/2, and 129X1. Endocrinology 145, 3307-3323. 575
Halaas, J. L., Gajiwala, K. S., Maffei, M., Cohen, S. L., Chait, B. T., Rabinowitz, D., 576
Lallone, R. L., Burley, S. K. and Friedman, J. M. (1995). Weight-Reducing Effects of the 577
Plasma-Protein Encoded by the Obese Gene. Science 269, 543-546. 578
Johansson, P., Bernstrom, J., Gorman, T., Oster, L., Backstrom, S., Schweikart, F., 579
Xu, B., Xue, Y. and Schiavone, L. H. (2013). FAM3B PANDER and FAM3C ILEI represent a 580
distinct class of signaling molecules with a non-cytokine-like fold. Structure 21, 306-13. 581
Kahle, M., Horsch, M., Fridrich, B., Seelig, A., Schultheiss, J., Leonhardt, J., Irmler, 582
M., Beckers, J., Rathkolb, B., Wolf, E. et al. (2013). Phenotypic comparison of common 583
mouse strains developing high-fat diet-induced hepatosteatosis. Mol Metab 2, 435-46. 584
Krycer, J. R., Sharpe, L. J., Luu, W. and Brown, A. J. (2010). The Akt-SREBP nexus: 585
cell signaling meets lipid metabolism. Trends Endocrinol Metab 21, 268-76. 586
Lahsnig, C., Mikula, M., Petz, M., Zulehner, G., Schneller, D., van Zijl, F., Huber, 587
H., Csiszar, A., Beug, H. and Mikulits, W. (2009). ILEI requires oncogenic Ras for the 588
epithelial to mesenchymal transition of hepatocytes and liver carcinoma progression. Oncogene 589
28, 638-50. 590
Li, J., Chi, Y., Wang, C., Wu, J., Yang, H., Zhang, D., Zhu, Y., Wang, N., Yang, J. 591
and Guan, Y. (2011). Pancreatic-derived factor promotes lipogenesis in the mouse liver: role of 592
the Forkhead box 1 signaling pathway. Hepatology 53, 1906-16. 593
Li, Z., Mou, H., Wang, T., Xue, J., Deng, B., Qian, L., Zhou, Y., Gong, W., Wang, J. 594
M., Wu, G. et al. (2013). A non-secretory form of FAM3B promotes invasion and metastasis of 595
human colon cancer cells by upregulating Slug expression. Cancer Lett 328, 278-84. 596
Pelleymounter, M. A., Cullen, M. J., Baker, M. B., Hecht, R., Winters, D., Boone, T. 597
and Collins, F. (1995). Effects of the obese gene product on body weight regulation in ob/ob 598
mice. Science 269, 540-3. 599
Robert-Cooperman, C. E., Carnegie, J. R., Wilson, C. G., Yang, J., Cook, J. R., Wu, 600
J., Young, R. A., Wolf, B. A. and Burkhardt, B. R. (2010). Targeted disruption of pancreatic-601
derived factor (PANDER, FAM3B) impairs pancreatic beta-cell function. Diabetes 59, 2209-18. 602
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 25
25
Robert-Cooperman, C. E., Dougan, G. C., Moak, S. L., Athanason, M. G., Kuehl, M. 603
N., Bell-Temin, H., Stevens, S. M., Jr. and Burkhardt, B. R. (2014). PANDER transgenic 604
mice display fasting hyperglycemia and hepatic insulin resistance. J Endocrinol 220, 219-31. 605
Robert-Cooperman, C. E., Wilson, C. G. and Burkhardt, B. R. (2011). PANDER KO 606
mice on high-fat diet are glucose intolerant yet resistant to fasting hyperglycemia and 607
hyperinsulinemia. FEBS Lett 585, 1345-9. 608
Song, Q., Sheng, W., Zhang, X., Jiao, S. and Li, F. (2013). ILEI drives epithelial to 609
mesenchymal transition and metastatic progression in the lung cancer cell line A549. Tumour 610
Biol. 611
Toye, A. A., Lippiat, J. D., Proks, P., Shimomura, K., Bentley, L., Hugill, A., Mijat, 612
V., Goldsworthy, M., Moir, L., Haynes, A. et al. (2005). A genetic and physiological study of 613
impaired glucose homeostasis control in C57BL/6J mice. Diabetologia 48, 675-686. 614
Waerner, T., Alacakaptan, M., Tamir, I., Oberauer, R., Gal, A., Brabletz, T., 615
Schreiber, M., Jechlinger, M. and Beug, H. (2006). ILEI: a cytokine essential for EMT, tumor 616
formation, and late events in metastasis in epithelial cells. Cancer Cell 10, 227-39. 617
Wang, C., Burkhardt, B. R., Guan, Y. and Yang, J. (2012). Role of pancreatic-derived 618
factor in type 2 diabetes: evidence from pancreatic beta cells and liver. Nutr Rev 70, 100-6. 619
Wang, O., Cai, K., Pang, S., Wang, T., Qi, D., Zhu, Q., Ni, Z. and Le, Y. (2008). 620
Mechanisms of glucose-induced expression of pancreatic-derived factor in pancreatic beta-cells. 621
Endocrinology 149, 672-80. 622
Wilson, C. G., Robert-Cooperman, C. E. and Burkhardt, B. R. (2011). PANcreatic-623
DERived factor: Novel hormone PANDERing to glucose regulation. FEBS Lett. 624
Wilson, C. G., Schupp, M., Burkhardt, B. R., Wu, J., Young, R. A. and Wolf, B. A. 625
(2010). Liver-specific overexpression of pancreatic-derived factor (PANDER) induces fasting 626
hyperglycemia in mice. Endocrinology 151, 5174-84. 627
Xu, W., Gao, Z., Wu, J. and Wolf, B. A. (2005). Interferon-gamma-induced regulation 628
of the pancreatic derived cytokine FAM3B in islets and insulin-secreting betaTC3 cells. Mol Cell 629
Endocrinol 240, 74-81. 630
Yang, J. and Guan, Y. (2013). Family with sequence similarity 3 gene family and 631
nonalcoholic fatty liver disease. J Gastroenterol Hepatol 28 Suppl 1, 105-11. 632
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 26
26
Yang, J., Robert, C. E., Burkhardt, B. R., Young, R. A., Wu, J., Gao, Z. and Wolf, 633
B. A. (2005). Mechanisms of glucose-induced secretion of pancreatic-derived factor (PANDER 634
or FAM3B) in pancreatic beta-cells. Diabetes 54, 3217-28. 635
Zhu, Y., Xu, G., Patel, A., McLaughlin, M. M., Silverman, C., Knecht, K., Sweitzer, 636
S., Li, X., McDonnell, P., Mirabile, R. et al. (2002). Cloning, expression, and initial 637
characterization of a novel cytokine-like gene family. Genomics 80, 144-50. 638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 27
27
FIGURE LEGENDS 659
Figure 1. Long and short-term fasting glycemic measurements of PANKO-C57. Mice of 660
both genders were 4 months of age. (A) Long-term fasting (overnight, 16 hours) blood glucose 661
measurements of male PANKO-C57 and WT mice (n= 12-15). (B) Short-term fasting (4 hours) 662
blood glucose measurements of males. (n= 8-11). Values are expressed as means ± SE. (C) 663
Long-term fasting blood glucose measurement of female PANKO-C57 and WT mice (n = 7-12). 664
(D) Short-term fasting (4 hours) blood glucose measurements of females. (n = 6-10). * P < 0.05, 665
**P < 0.01 by Student t-test. 666
667
Figure 2. Enhanced glucose tolerance in male PANKO-C57 mice. All mice described below 668
were 4 months of age. (A) Intraperitoneal (i.p.) glucose tolerance tests (GTTs) were performed 669
following overnight fast (approximately 16 hours) on male PANKO-C57 and WT mice by 670
injection of glucose at 2 g/kg and measurement of plasma glucose concentration at the indicated 671
time points (n= 8-12). (B) Intraperitoneal insulin tolerance tests (ITTs) were performed on male 672
PANKO-C57 and WT mice following a 4-hour fast by injection of insulin at 1 unit/kg and 673
subsequent measurement of plasma glucose concentration at all indicated time points. Results are 674
presented as the percentage of baseline glucose concentration measured at time point 0 (n= 8-675
11). (C) GTT performed on female PANKO-C57 and WT mice as described above (n= 7-12). 676
(D) ITT performed on female PANKO-C57 and WT mice as described above (n= 7-11). Values 677
are expressed as the mean + SE. *P < 0.05, **P < 0.01 as determined by two-way ANOVA. 678
679
Figure 3. Increased body weight of male and female PANKO-C57. Body weight 680
measurements were recorded at similar times from 8 to 23 weeks of age following an overnight 681
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 28
28
fast. (A) Male PANKO-C57 and WT mice weight over time (n = 18-27). (B) Female PANKO-682
C57 and WT mice weight over time (n= 13-22). Values are expressed as the means + SE. *P < 683
0.05, **P < 0.01, ***P < 0.001 as determined by two-way ANOVA. 684
685
Figure 4. Fasting hormonal evaluation of male PANKO-C57. Fasting plasma levels of (A) 686
insulin, (B) glucagon and (C) leptin were measured at 2 and 5 months of age in male PANKO-687
C57 and WT mice plasma fasted overnight using Milliplex Mouse Metabolic Hormone Magnetic 688
Bead Panel (Millipore) for MAGPIX® Luminex system (n= 4-6). (D) Overnight fasting levels of 689
C-Peptide from mice aged 2 months using the MAGPIX® Luminex system as above (n= 4). (E) 690
Corticosterone levels were measured from plasma collected during fasting conditions at 2 and 5 691
months of age using the Corticosterone EIA Kit (Enzo Life Sciences) (n= 3). (F) Plasma insulin 692
levels during course of GTT (n= 3-5). Values are expressed as means ± SE. *P < 0.05, **P < 693
0.01, ***P < 0.001 as determined by Student t-test for figures 4A-4E. Two-way ANOVA was 694
performed for Figure 4F. 695
696
Figure 5. Hepatic glycogen, triglyceride and gluconeogenic content and expression in 697
PANKO-C57 698
(A) Hepatic glycogen content (in μg/μl of tissue lysate) measurement of male PANKO-C57 and 699
WT mice at 4-5 months of age following insulin stimulation (2U/kg) or an overnight fast using 700
Glycogen Assay Kit (Abcam®) (n= 3 per group). (B) Hepatic triglyceride content evaluation of 701
4-5 month old male PANKO-C57 and WT mice as described in A, using 100 mg of liver tissue 702
and Triglyceride Quantification Kit (Abcam®) (n= 3). (C) Serum triglycerides from fasting 703
serum of PANKO-C57 and WT male mice were analyzed longitudinally at 2 and 5 months of 704
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 29
29
age using the Triglyceride Quantification Kit (Abcam®) (n= 5). (n=4). Values are expressed as 705
the mean + SE. *P < 0.05, **P < 0.01 as determined by Student t-test. 706
707
Figure 6. Western analysis of PANKO-C57 hepatic signaling. (A) Western blot analysis of 708
liver lysate from male PANKO-C57 and WT mice fasted overnight for protein levels of p-PI3K 709
(Tyr 508), total and p-Akt (Thr 308), mature and precursor SREBP-1, total and p-AMPKα (Thr 710
172), and GAPDH serving as loading control. Lanes 1-3 correspond to PANKO-C57 mice and 711
lanes 4-6 correspond to age and gender matched C57BL/6J WT mice. * denotes statistical 712
significance of P < 0.05 as determined by measurement with ImageJ as detailed in B-E. (B) 713
Densitometric analysis of hepatic protein levels of phosphorylated proteins PI3K, (C) Akt, (D) 714
AMPK, and mature levels of SREBP-1 were determined and normalized to total levels of 715
respective protein followed by total protein normalization to GAPDH (n= 3). Values are 716
expressed as the mean + SE. *P < 0.05 as determined by unpaired Student t-test. 717
718
719
720
721
722
723
724
725
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 30
30
RESOURCE IMPACT 726
Background 727
PANcreatic-DERived factor (PANDER, FAM3B) was originally cloned in 2002 and is a 728
member of the superfamily of FAM3 genes. PANDER appears to serve a role in regulation of 729
glycemic levels, insulin action and hepatic lipogenesis. Overexpression of PANDER results in a 730
phenomenon known as selective hepatic insulin resistance whereby insulin signaling is blunted, 731
yet lipogenesis is increased. SHIR is a hallmark pathogenic paradox of type 2 diabetes (T2D), 732
which is the most common global metabolic disorder with ever increasing rates. Despite 733
numerous review articles eluding to PANDER serving a potential role in the onset or progression 734
of T2D, no stable animal models have been generated on well-established genetic backgrounds 735
of T2D susceptibility. Therefore, there is a strong need for novel animal models on congenic 736
backgrounds with discernible phenotypes for the investigation of PANDER. 737
Results 738
In this study, the authors have generated a PANDER knockout mouse model on a pure C57BL/6 739
background (PANKO-C57) to promote the phenotypic penetrance of PANDER and provide a 740
useful tool for subsequent studies. In contrast to the prior PANDER knockout model, the 741
PANKO-C57 model exhibited increased body weight, enhanced glucose tolerance during both 742
fed and fasting conditions. This phenotype was more significant in PANKO-C57 males versus 743
female mice, however females still displayed overall trends in enhanced glucose tolerance. In 744
addition, fasting plasma insulin and c-peptide levels were concordantly decreased in the 745
PANKO-C57 mouse along with increased leptin levels. Hepatic insulin signaling was 746
significantly increased during fasting conditions as demonstrated by increased phosphorylation 747
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 31
31
of hepatic Akt and AMPK along with mature SREBP-1 expression in the PANKO-C57 model. 748
Insulin stimulation of PANKO-C57 mice resulted in increased hepatic triglyceride and glycogen 749
content as compared to C57BL/6J WT. Altogether, this model has enhanced the understanding 750
regarding the mechanistic implications of PANDER and has unified many of the prior conflicted 751
findings regarding the role in glycemic regulation. 752
Implications and future directions 753
The study provides evidence that this generated PANDER animal model derived on a well-754
established background for T2D displays an enhanced phenotype that can be easily employed in 755
future investigations. Given the limited availability of current PANDER animal models, the 756
PANKO-C57 suitably fills this critical gap. The PANKO-C57 has a strong breeding capacity 757
and discernible phenotype to allow for the future creation of additional animal models and 758
investigations to evaluate the role of PANDER in both T2D and glycemic regulation. 759
760
761
Dise
ase
Mod
els &
Mec
hani
sms
D
MM
Acce
pted
man
uscr
ipt
Page 32
Figure 1. Moak et al.
A. B.
WT KO0
50
100
150
200
250
Blo
od G
luco
se (
mg/
dl)
C. D.
Page 33
Figure 2. Moak et al. REVISION
A. B.
C. D.
Page 34
Figure 3. Moak et al. REVISED
Page 35
Figure 4. Moak et al. REVISION
A. B. C.
D. E.
F.
KO
0.1
0.2
0.3
0 15 30 60
WT
120
*
Time after IP glucose injection (min)
Insu
lin
(n
g/m
l)
Page 36
Figure 5. Moak et al.
A. B.
C.
*
Page 37
PANKO-C57 C57BL/6J WT
P-PI3K
Total Akt
P-Akt
Mature SREBP-1c
Pre-SREBP-1c
GAPDH
Total AMPKα
P-AMPKα
Figure 6. Moak et al.
*
*
*
*
A.
D.
E.
B.
E.
C.