Transcriptome Profiles and Gene Expression of Min 6 Cells Are Altered by Pancreatic Stellate Cells

Aim: To identify the influence of pancreatic stellate cell (PSCs) secretions on gene expression profiles of Min6 cells by whole transcriptome sequencing. Methods: Pancreatic stellate cells (PSCs) were isolated from C57BL6J mice and propagated in vitro to acquire the activated phenotype. Total RNA was isolated from monocultured (MC) and PSC cocultured (CC) Min6 cells to prepare cDNA libraries, which were subjected to whole transcriptome sequencing for identifying differential expression of β-cell transcription factors (Pdx-1, Rfx6 and NeuroD1) related to insulin gene transcription and GSIS related genes such as Glut2, Gck, Abcc8, Kcnj11 and L-type Ca channels (Cacnb2, Cacna1c). qRT-PCR was used to validate the gene expression. GSIS of Min6 cells was examined by estimating insulin levels in response to high glucose challenge. Results: Transcriptome analysis of discovery set revealed that coculture of Min6 cells with PSCs caused increased expression of β-cell specific genes (Ins1, Rfx6 and NeuroD1) concomitant with decreased expression of Pdx-1, MafA and Nkx2-2. Expression of GSIS associated genes (Glut2, Gck, Abcc8, Kcnj11 and Cacnb2) was decreased in such conditions. Validation by qRT-PCR in Min6 cells cocultured with PSCs revealed increased significant expression of Ins1 (2.1 ± 0.22 folds; p ≤ 0.001), Rfx6 (1.68 ± 0.23 folds; p ≤ 0.002) and NeuroD1 (0.96 ± 0.11 folds; p ≤ 0.01), accompanied by downregulation of Cacnb2 (−0.93 ± 0.57 folds; p ≤ 0.05). PSC secretions did not restore the GSIS from glucose unresponsive higher passage Min6 cells (MC: 1.33 ± 0.42; CC: 1.55 ± 0.72 pmol/mg protein; p = ns) upon high glucose stimulation. However, glucose responsive higher passage Min6 cells cocultured with PSCs presented increased insulin secretion (MC: 7.025 ± 0.64; CC: How to cite this paper: Bynigeri, R.R., Mitnala, S., Vishnubhotla, R., Talukdar, R., Singh, S.S. and Duvvuru, N.R. (2018) Transcriptome Profiles and Gene Expression of Min6 Cells Are Altered by Pancreatic Stellate Cells. Advances in Bioscience and Biotechnology, 9, 273-288. https://doi.org/10.4236/abb.2018.96018 Received: April 26, 2018 Accepted: June 25, 2018 Published: June 28, 2018 Copyright © 2018 by authors and Scientific Research Publishing Inc. This work is licensed under the Creative Commons Attribution International License (CC BY 4.0). http://creativecommons.org/licenses/by/4.0/ Open Access R. R. Bynigeri et al. DOI: 10.4236/abb.2018.96018 274 Advances in Bioscience and Biotechnology 14.84 ± 1.01 pmol/mg protein; p ≤ 0.04) concomitant with marginal increase of insulin contents. Conclusion: PSC secretions increase Ins1, Rfx6 and NeuroD1 gene expression, GSIS from glucose responsive Min6 cells, but do not restore the GSIS from glucose unresponsive Min6 cells.

Recent studies with rodents involving coculture of activated PSCs with islets and with β cell lines indicated variable influence of PSCs on insulin secretory response of β-cells.Rat insulinoma cell lines (RINm-5F, INS-1) showed decreased insulin secretion [14] [32] when cocultured with PSCs or incubated with PSC conditioned medium.In yet another study, increased insulin secretory response associated with a decrease in total insulin content was reported when mouse islets were cocultured with PSCs [33].Availability of adequate information regarding gene expression in β-cells in response to the presence of PSCs in coculture conditions, would resolve these observations and comprehensively elucidate the influence of PSCs on β-cell functions.In view of these considerations, the present study involving indirect coculture of Min6 cells with activated PSCs was aimed to 1) study the transcriptome profile alterations and the accompanying 2) expression of β-cell specific transcription factors and GSIS related genes in Min6 cells.

Chemical
All the chemicals used in the experiments are obtained from Sigma unless otherwise mentioned.

Mouse insulinoma cells
Min6 cells were obtained from National Centre for Cell Sciences, Pune, India as well as from AddexBio, San Diego, USA.Min6 cells were routinely cultured in DMEM containing 25 mM glucose supplemented with 15% FBS in T25 flasks and maintained in a humidified chamber with 5% CO 2 at 37˚C.Where required, Min6 cells at lower passage (P11) were cultured to higher passages (P53-64) under the above conditions.

Isolation and activation of pancreatic stellate cells
Male C57BL6J mice (approximately three months old) were purchased from National Centre for Laboratory Animal Sciences, Hyderabad and were used to isolate the quiescent stellate cells from mouse pancreas as per the method described earlier [34], with minor modifications.Briefly, surgically resected mouse pancreatic tissue was inflated by injecting Hank's balanced salt solution (HBSS) containing Collagenase P (Roche, Germany), Protease type IV and DNase I (Roche, Germany).The distended tissue was incubated in a shaking water bath at 37˚C for 5 -7 minutes.Finely minced tissue was further incubated at 37˚C for 5 minutes in a shaking water bath.The digested tissue was then centrifuged and pelleted at 450 ×g for 10 minutes.The resultant supernatant was discarded and the pellet was resuspended in HBSS containing 0.3% BSA and centrifuged.The obtained pellet was resuspended in 10 mL of 0.3% BSA containing HBSS and 8 mL of Nycodenz® (Axis Shield, Norway) gradient solution.The cell suspension was carefully layered under HBSS buffer containing 0.3% BSA and centrifuged at 1400 ×g for 20 minutes at 4˚C.The fuzzy band formed at the interphase was collected and centrifuged at 450 ×g for 10 minutes.The obtained pellet with primary cells was suspended in IMDM (Himedia, India) containing 20% FBS (Himedia, India) along with 1% penicillin and streptomycin (Himedia, India).
The quiescent phenotype of isolated PSCs was ascertained by appearance of lipid droplets within 6 -8 hours.These primary cells are subsequently passaged to obtain the activated phenotype of PSCs under in vitro conditions.All the indirect coculture experiments are conducted using activated PSCs of passage 3.

Indirect Coculture of PSCs with Min6 Cells
Min6 cells were subjected to indirect coculture with activated PSCs using transwell inserts with polycarbonate membrane (Himedia, India) having a pore size of 0.4 µm as per the method described earlier [14] [35].PSCs (25 × 10 3 cells) suspended in complete IMDM were seeded on a transwell insert followed by seeding of Min6 cells (0.25 × 10 6 cells) suspended in complete DMEM into the culture well plate.After separate incubation of cells for 24 -36 hours, the spent media were replaced with fresh media and culture inserts with PSCs were placed in culture wells containing Min6 cells.Incubation was continued for the next 72 hours followed by harvesting Min6 cells for further studies involving RNA isolation, transcriptome profiling and qRT-PCR validation of the target gene expression as well as GSIS assays.

Whole Transcriptome Sequencing
Total RNA was isolated from Min6 cells employing RNeasy kit (Qiagen, Germany) as per the manufacturer's protocol.Quantity and quality of the isolated RNA was assessed using Bioanalyzer 2100 (Agilent, Germany) and samples which showed RNA integrity number > 8 were used in this study.Whole transcriptome sequencing was performed on monocultured and PSC cocultured Min6 cells samples employing Ion Total RNA-Seq kit V2, including Ion RNA-Seq Core kit v2, Ion RNA-Seq Primer Set (Life technologies, USA).Briefly, 5 µg of total RNA was purified using RiboMinus™ Eukaryote System v2 Kit to obtain rRNA-depleted total RNA.RNA thus obtained was fragmented using RNase III for 10 minutes and purified using magnetic beads.Purified RNA was quantitated on Agilent Bioanalyzer using RNA 6000 Pico Kit which yielded fragments in the range of 100 -200 nucleotides.Whole transcriptome libraries were constructed by hybridizing and ligating the RNA and subjected to reverse transcription.The obtained cDNA was purified and amplified by polymerase chain reaction.The yield and size distribution of the amplified cDNA was assessed; barcoded pooled transcriptome libraries were sequenced on Proton Semiconductor Sequencer (Ion torrent, Life Technologies, USA).Whole transcriptome sequencing was performed using a single set of total RNA samples isolated from monocultured and PSC cocultured Min6 cells.Sequencing reads were assessed for quality and single-end mRNA reads were mapped onto Mouse genome (mm10) and analyzed using Partek Genomics Suite v6.6.BAM files were imported into the software; samples were assigned the respective attributes using categorical (Coculture Vs Monoculture) variables.

Validation of target gene expression by qRT-PCR
Total RNA was isolated from both monocultured and PSC cocultured Min6 cells using TRIzol® reagent (Ambion®, Life Technologies, USA) as per the manufacturer's protocol.1 µg of total RNA was used to synthesize the cDNA reaction using Superscript IV First-Strand Synthesis System (Invitrogen, Lithuania) in a total volume of 20 μL reaction setup.The prepared cDNA was used to validate the differential expression of β-cell specific and GSIS related genes that were identified in the discovery study, using  expression was normalized to the expression levels of β-actin.The fold difference between the monoculture and coculture samples was calculated by using Pfaffl's 2 −ΔΔCt method and the obtained values were log2 transformed.

Glucose Stimulated Insulin Secretion
Glucose Stimulated Insulin Secretion (GSIS) from Min6 cells was examined as per the method described earlier [24]

Measurement of Insulin Contents in Min6 Cells
Min6 cells were lysed in Radio Immuno Precipitation Assay (RIPA) buffer (Cell Signaling Technologies, USA) and incubated on ice for 5 minutes.The cell lysates were then briefly sonicated and centrifuged at 14,000 ×g for 10 minutes at 4˚C.The concentration of the total protein in these samples was measured using Bradford's method.The supernatants were collected and stored at −20˚C until further use to measure the total insulin content.Insulin was estimated using mouse insulin ELISA (Mercodia, Sweden) kit and OD was recorded at 450 nm on a microplate reader (BioRad Model 680, Japan).Insulin contents were normalized to total protein and expressed in terms of pmoles of insulin per mg protein.

Statistical Analysis
Probability (p) values between the MC and CC groups were calculated using Student's t-test using Microsoft Excel program.Data are represented as mean +/− SEM.p ≤ 0.05 is considered to be statistically significant.

Whole Transcriptome Analysis Identifies Differentially Expressed Genes in Higher Passage Min6 Cells Cocultured with PSCs
The discovery set of cDNA libraries, with an average size of 248 and 269 bp length respectively, were prepared from Min6 cells cultured in presence or in absence of PSCs and subjected to whole transcriptome sequencing.These cDNA libraries generated ≈ 67 million reads with a mean length of 107 bp and 54% usable reads for MC higher passage Min6 cells and ≈58 million reads with 92 bp as mean length and 46% usable reads for CC samples.The usable reads obtained from the whole transcriptome sequencing were aligned separately on to the mouse mm10 reference genome, which identified 29610 transcripts during the analysis.Expression levels of the transcripts were normalized as reads per kilobase per million mapped reads (RPKM) to identify the differentially expressed genes.Of all the genes subjected to transcriptome analysis in the discovery set, important changes could be noted with regard to increased expression of β-cell specific genes such Ins1 (Figure 1(a)).

PSCs Do Not Alter the Expression of GSIS Related Genes in Higher Passage Min6 Cells
As in case of β-cell related genes, changes in the expression levels of GSIS associated genes are also validated by qRT-PCR.Although the expression levels of Glut2 (−1.12 ± 0.76 folds; p ≤ 0.76), potassium channel subunit Abcc8 (−0.61 ± 0.38 folds; p ≤ 0.18) and L-type calcium channel subunit Cacnb2 (−0.93 ± 0.57 folds; p ≤ 0.05) were decreased as in the discovery set, results of the validation revealed no statistically significant change in the expression levels between mono and PSC cocultured Min6 cells.Although not significant, marginal increase in the expression of Gck (0.63 ± 0.81 folds; p ≤ 0.48) and potassium channel subunit Kcnj11 (0.53 ± 0.57 folds; p ≤ 0.40) were noticed in PSC cocultured Min6 cells compared to monocultured Min6 cells.The L-type calcium channel, Cac-na1c (0.60 ± 0.38 folds; p ≤ 0.38) also did not show any significant increase in its

PSC Secretions Do Not Restore GSIS from Higher Passage Glucose Unresponsive Min6 Cells
GSIS response of Min6 cells at high passage (P53-64) was examined when they were cultured in presence or in absence of PSC secretions and the obtained results are depicted in Figure 2

Discussion
The present communication describes results obtained upon culturing Min6 cells exposed to secretions of pancreatic stellate cells.Although PSC infiltration into the islets was demonstrated earlier in pancreatic disease including chronic pancreatitis and pancreatic cancer as well as in diabetic rodent models [14] [15] Ins1 gene expression were unaccompanied by changes in genes related to insulin secretory pathway.The enhanced expression of Rfx6 and NeuroD1 suggests that Rfx6, either on its own or in combination with NeuroD1 [36], does influence expression of Ins1 gene.These results gain credence from observations made in an earlier study suggesting the importance of Rfx6 [37] [38], on β-cell insulin gene expression.
Despite the increased expression of Rfx6, which influences GSIS related changes [37] [38], our studies did not reveal altered expression of GSIS associated genes in response to PSC secretions on Min6 cells.Results obtained with indirect coculture did not reveal changes in GSIS of Min6 cells exposed to PSCs.
Recent in vitro coculture studies presented varied results on influence of PSCs on GSIS response [13]- [18].Studies involving indirect cocultue of PSCs and its secretions with RINm-5F and INS-1 cell lines reported decreased insulin secretion [14] [32], while direct coculture of mouse islets in presence of PSCs caused increased insulin secretion [33].While considering the response of insulin secreting cells to PSCs under in vitro conditions, the innate ability of cells to respond to glucose challenges is of considerable importance.It is generally held that RINm-5F cells are not responsive to glucose challenges [39] [40], and the ability of Min6 cells to respond to glucose depends on the number of passages they undergo [41] [42].
In our study we employed Min6 cells of higher passage (P53-64) and examined their insulin secretory responses.These cells did not exhibit much change in their GSIS response either in monoculture or coculture.Such an observation may be ascribed to the reported inability of glucose unresponsive higher passage Min6 cells to high glucose challenge [41] [42].This result also indicates the inability of PSC secretions to restore the lost ability of insulin secretion of higher passage Min6 cells to glucose challenge.In order to verify such an inference, we examined GSIS of lower passage Min6 cells (P11) that were cultured to higher passage (P53-64).Results of this experiment as shown in Figure 3(a) demonstrated the ability of such cells not only to respond to glucose challenge but also confirmed that PSC secretions further enhance insulin release by Min6 cells.The ability of higher passage Min6 cells to respond to glucose challenge in presence of PSC secretions also denotes that the glucose responsive nature of Min6 cells is of key relevance for the PSCs to influence insulin secretory abilities of Min6 cells.It is apparent that results obtained upon in vitro studies might not reflect in vivo situation, more so under disease conditions.Extrapolating the results reported herein to a physiological context, we may infer that even though PSCs can innately influence insulin secretory abilities of β-cells, it is possible that such an ability might be diminished in pancreatic disease.
The nature of secretions made by PSCs has not been characterized in the present study.It is known that PSC secretions, including inflammatory cytokines and different growth factors, can not only influence the pancreatic milieu but also influence β-cell functions.Importantly, the nature of PSC secretions under in vivo disease conditions would largely determine their influence on β-cell functions.Despite these limitations, this is the first report on the influence of PSCs on transcriptome profiles as well as genes that are involved in the insulin synthesis and secretory pathway of higher passage Min6 cells.Further experiments including characterization of PSC secretome should yield important information about their influence on gene expression profiles and insulin secretory response from islet isolates obtained from diabetic conditions.

expression (Figure 1
(b)), which was noticed to be upregulated by 1.2-fold in the discovery study.Advances in Bioscience and Biotechnology

Figure 1 .
Figure 1.Relative expression of β-cell specific and GSIS related genes in PSC cocultured higher passage Min6 cells: (a) PSC cocultured Min6 cells showing a significant increase in the expression levels of Ins1, Rfx6 and NeuroD1, while Pdx1, MafA and Nkx2-2 expression levels were seen unaltered; (b) GSIS associated genes such as Glut2, Gck, Abcc8, Kcnj11 and Cacna1c showing no significant change in their expression levels while Cacnb2, showed significant downregulation in PSC cocultured Min6 cells (n = 5).β-actin was used as an endogenous control to normalize the gene expression levels.Fold changes were calculated using 2 −ΔΔCt method and values were log transformed.Each target gene studied was analysed in duplicates.Data are represented as mean ± SEM. *p ≤ 0.001, **p ≤ 0.002, ***p ≤ 0.01, ****p ≤ 0.05.MC = Monoculture, CC = Coculture.
(a).It was noted that the high passage Min6 cells used in this study are unresponsive to high (25 mM) glucose stimulation and coculturing such cells with of PSC (MC: 1.33 ± 0.42; CC: 1.55 ± 0.72 pmol/mg protein; p = ns) did not reveal any statistically significant changes (Figure2(a)).However, repeated subculturing of Min6 cells from passage 11 to passage 53-64 resulted in enhanced glucose responsiveness (MC: 7.025 ± 0.64; CC: 14.84 ± 1.01 pmol/mg protein; p ≤ 0.04) of Min6 cells when cocultured with PSCs compared to monocultured Min6 cells (Figure3(a)), suggesting that the influence of PSC secretions is dependent on the ability of Min6 cells to respond to high glucose stimulation.As shown in Figure2(b), the total insulin contents of glucose unresponsive higher passage Min6 cells were found to be 251.31± 53.95 and 296.37 ± 56.75 pmol/mg protein (p ≤ 0.59), whereas the glucose responsive higher passage Min6 cells showed 4871.94 ± 1271.23 and 6454.03 ± 474.21 pmol/mg protein (p ≤ 0.51) (Figure 3(b)) in respective monocultured and PSC cocultured Min6 cells.

Figure 2 .Figure 3 .
Figure 2. Impaired GSIS response from higher passage Min6 cells: (a) Mono and PSC cocultured Min6 cells showing no significant change in the amount of insulin secreted when subjected to high (25 mM) glucose stimulation (n = 4), compared to basal (2.5 mM) glucose, suggesting the impaired glucose sensing ability of these higher passage Min6 cells; (b) PSC cocultured Min6 cells showing a marginal but not significant increase in the total insulin contents compared to monocultured Min6 cells (n = 3).Insulin secretion and insulin content was normalized to protein levels.Each sample was quantitated in duplicates and data are represented as mean ± SEM. p = ns.GSIS = Glucose Stimulated Insulin Secretion.Glc = Glucose.

[ 30 ]Figure 4 .
Figure 4. Relative expression of β-cell specific and GSIS related genes in PSC cocultured glucose responsive higher passage Min6 cells: (a) PSC cocultured glucose responsive higher passage Min6 cells showing increased expression of Ins1, Rfx6 and NeuroD1 with no significant changes in the expression of Pdx1, MafA and Nkx2-2, similar to PSC cocultured glucose unresponsive higher passage Min6 cells (n = 3); (b) Significant upregulation of Cacnb2 and Cacna1c calcium channels with no alteration of other studied GSIS genes in cocultured glucose responsive higher passage Min6 cells, confirms that PSC secreted factors influence the expression of β-cell specific and GSIS associated genes in higher passage Min6 cells (n = 3).β actin was used as an internal control to normalize the target gene expression and obtained fold changes were log transformed.Each target gene studied was analyzed in duplicates.Data are represented as mean ± SEM. *p ≤ 0.008, **p ≤ 0.05, ***p ≤ 0.005 Vs monocultured cells.

Table 1 .
Primer sequences used in qRT-PCR study.
the Power Sybr® green PCR Master Mix (Applied Biosystems, United Kingdom) on Step One® Real Time PCR System (Applied Biosystems, Singapore).Forward and reverse primer sequences for the target genes used in this study are listed in Table 1.Relative target gene Advances in Bioscience and Biotechnology hours, the cells were washed twice in KRBH buffer and pre-incubated in KRBH containing 2.5 mM glucose for 30 minutes.The cells were incubated for one hour in presence of 2.5 mM (basal) glucose, followed by stimulation with 25 mM (high) glucose, for another hour.The cell supernatants collected after basal and high glucose stimulation were stored at −20˚C until further use.

Table 2 .
List of differentially expressed β-cell specific and GSIS related genes in PSC cocultured higher passage Min6 cells identified in whole transcriptome analysis from the discovery set.Reads, Reads Per Kilobase Million (RPKM) values and the respective fold changes for the mentioned genes were obtained after analysis of the data using Partek Genomics Suite v6.6.

qRT-PCR Validation Identifies Increased Expression of Ins1, Rfx-6 and NeuroD1 Genes in PSC Cocultured Higher Passage Min6 Cells
Changes in the expression levels of β-cell related genes observed in the discovery set were validated by qRT-PCR.In comparison to Min6 cells cultured in absence of PSCs, those cultured in presence of PSCs revealed increased expression levels