Vil-Cre specific Schlafen 3 knockout mice exhibit sex-specific differences in intestinal differentiation markers and Schlafen family members expression levels

The intestinal epithelium requires self-renewal and differentiation in order to function and adapt to pathological diseases such as inflammatory bowel disease, short gut syndrome, and ulcers. The rodent Slfn3 protein and the human Slfn12 analog are known to regulate intestinal epithelial differentiation. Previous work utilizing a pan-Slfn3 knockout (KO) mouse model revealed sex-dependent gene expression disturbances in intestinal differentiation markers, metabolic pathways, Slfn family member mRNA expression, adaptive immune cell proliferation/functioning genes, and phenotypically less weight gain and sex-dependent changes in villus length and crypt depth. We have now created a Vil-Cre specific Slfn3KO (VC-Slfn3KO) mouse to further evaluate its role in intestinal differentiation. There were increases in Slfn1, Slfn2, Slfn4, and Slfn8 and decreases in Slfn5 and Slfn9 mRNA expression that were intestinal region and sex-specific. Differentiation markers, sucrase isomaltase (SI), villin 1, and dipeptidyl peptidase 4 and glucose transporters, glucose transporter 1 (Glut1), Glut2, and sodium glucose transporter 1 (SGLT1), were increased in expression in VC-Slfn3KO mice based on intestinal region and were also highly female sex-biased, except for SI in the ileum was also increased for male VC-Slfn3KO mice and SGLT1 was decreased for both sexes. Overall, the variations that we observed in these VC-Slfn3KO mice indicate a complex regulation of intestinal gene expression that is sex-dependent.


Introduction
The murine Schlafen protein family is organized into three groups based on their molecular mass and structure. Slfn1 and -2 are the shortest and are in Group I. Slfn3 and -4 are in Group II and are intermediate in mass. The longest Slfns are in Group III and included Slfn5, -8, -9, and -14 [1,2]. All of the Slfns share a specific slfn box domain that is next to a divergent AAA domain [2,3]. However, a SWADL (Ser-Trp-Ala-Asp-Leu) domain appears only in the Group II and III Slfn family members. Finally, only Group III Slfns contain a C-terminal nuclear targeting sequence (RKRRR) that is homologous to DNA/RNA helicase superfamily I, and so Slfns in Groups I and II are cytosolic proteins since they do not contain this C-terminal nuclear targeting sequence [2][3][4][5][6]. The human Slfn proteins include SLFN5, SLFN11, SLFN12, SLFN13, and SLFN14 [7]. SLFN12 is the only protein that remains in the cytoplasm and is most closely homologous to murine Slfn3 and Slfn4 from Group II [7]. The most prominent role of this gene family is their control of cell differentiation, cell growth, and cell cycle progression [8]. Much work has also focused on the role of Slfn3, 4, 5, and 8 in the regulation of T cell differentiation, activation, and proliferation [2,9,10]. More specifically, Slfn3 is necessary for the regulation of intestinal differentiation, development, and maturation [11][12][13][14][15][16]. There is a direct correlation between the expression of Slfn3 and the expression of epithelial differentiation markers sucrase isomaltase (SI), villin1 (Vil1), dipeptidyl peptidase 4 (Dpp4), glucose transporter 2 (Glut2), and sodium glucose transporter 1 (SGLT1), which is dependent on the P-loop region in the N-terminus of Slfn3 [12,[15][16][17]. Previous work utilizing a pan-Slfn3 knockout (Slfn3KO) mouse demonstrated that Slfn3 affects the mRNA expression levels of Slfn family members in the intestinal mucosa, thymus, and spleen [18]. Additionally, the global loss of Slfn3 results in sex-dependent disturbances in metabolic pathways, genes related to adipogenesis and ketogenesis, and intestinal differentiation markers [16]. This was associated with less weight gain in the Slfn3KO mice in comparison to wild type (WT) mice, with the most striking decrease in female Slfn3KO mice [16]. We have further studies involving the Slfn3 human ortholog, SLFN12, where we demonstrated that SLFN12 promotes the expression of SI by binding to SerpinB12 through its ATP-binding region [19]. This binding of SLFN12 to SerpinB12 results in the stimulation of USP14 and UCHL5 deubiquitylase activity, leading to increases in Cdx2 levels. Determination of this pathway may allow for its targeting and manipulation in human enterocytic differentiation in order to treat mucosal atrophy, short gut syndrome, and possibly obesity [19].
The small intestinal mucosa is a heterogenous environment of immune cells and intestinal epithelial cells, and Slfn family members have functional roles in both cell populations. Therefore, we now have sought to specifically target Slfn3 in epithelial cells to better define its functional role in intestinal epithelial differentiation. We generated a Vil-Cre specific Slfn3 knockout mouse and evaluated changes in intestinal differentiation marker and Slfn family member expression.

Single-molecule RNA in situ hybridization
RNA in situ hybridization experiments were performed by Advanced Cell Diagnostics, Inc.
(Newark, CA) using the RNAscope1 technology, which has been previously described [20]. Paired double-Z oligonucleotide probe was designed against Slfn3 target RNA using custom software. The Probe Design # was NPR-0001265 and was named BA-Mm-Slfn3-2zz-st. This specific probe for Slfn3 targeted 560-1785 of NM_011409.1. FFPE tissue section samples were prepared according to manufacturer's recommendations. Each sample was quality controlled for RNA integrity with a probe specific to the housekeeping gene PPIB. Negative control background staining was evaluated using a probe specific to the bacterial dapB gene. Brightfield/ fluorescent images were acquired using a Leica-SS7225 microscope with a 40x objective.

Histology
Ileum was collected and was fixed in 10% formalin at room temperature. After 48 hours the tissue was transferred to 70% ethanol. Thereafter, tissues were processed using the Lynx II Tissue Processor according to the manufacturer's instructions, embedded in paraffin, sectioned, and stained with the assistance of the Histology core (Department of Biomedical Sciences, University of North Dakota). 5-8μm transverse sections of the intestines were stained with H&E for further analysis.

RNA isolation and qPCR
Intestinal mucosa was collected from 6-17 week old male and female Ctrl and VC-Slfn3KO mice. Total RNA was isolated from intestinal duodenum, jejunum, and ileum using the RNeasy Lipid Kit and the QiaCube instrument per manufacturers' protocols (Qiagen, Valencia, CA). Preparation of cDNA from RNA samples was performed using the SMARTScribe Reverse Transcription kit (Takara Clontech, Mountain View, CA). qPCR analysis of the cDNA samples were analyzed using the BioRad CFX96 Touch Real-Time PCR Detection System and the PrimeTime Gene Expression Master Mix from Integrated DNA Technology (IDT, Coralville, IA). RNA expression levels were ascertained from the threshold cycle (Ct) values using the 2 -ΔΔCt method using RPLP0 as the reference control gene. Primer/probe sets and qPCR cycle conditions were previously published [16].

Intestinal epithelial cell isolation & flow cytometry
Intestinal cells were isolated as previously described with a few modifications [21]. Briefly, intestinal regions were opened longitudinally and cut into 0.5 cm pieces and placed in a 50 ml

Statistics
Quantitative PCR was assessed by 2-way ANOVA with Uncorrected Fisher's LSD. All other data was compared by unpaired, two-tailed t-test. We sought 95% confidence in our statistical analysis and corrected for multiple tests as necessary. Data are represented as mean ± SE.

Generation of VC-Slfn3KO mice
Mice carrying a flexible gene-trap knockout-first, lacZ, tagged insertion allele of Slfn3 (Slfn3 βgeo mice) were created at the UC Davis KOMP Repository (Fig 1). These mice were further bred with B6.129S4-Gt(ROSA)26Sor tm2(FLP � )Sor /J (Flpo) mice and then B6.Cg-Tg(Vil1-cre) 1000Gum/J (VilCre) mice (Fig 1 and S1 Fig). Validation of Vil-Cre, intestinal epithelial cellspecific knock out of Slfn3 was confirmed by single-molecule RNA in situ hybridization (Fig  2A). VC-Slfn3KO mice have little to no Slfn3+ intestinal epithelial cells compared to Ctrl mice ( Fig 2B). Slfn3 + immune cells are seen within the lamina propria of the villi indicating the Vil-Cre specificity to the intestinal epithelial cells. The loss of Slfn3 in pan-Slfn3KO mice caused a decrease in ileal villus length in males, an increase in ileal villus length in female Slfn3KO mice, and a decrease in crypt depth in both male and female Slfn3KO mice [16]. We observed the same significant changes in villus length and crypt depth for the VC-Slfn3KO mice in both the male and female mice (Fig 3).

Slfn3 influences Slfn family member expression
Previously, we demonstrated that expression of Slfn family members are affected by the loss of Slfn3 in the ileum of pan-Slfn3KO mice [18]. The loss of Slfn3 in the intestinal epithelial cells of VC-Slfn3KO mice also affected the expression of Slfn family members in the duodenal, jejunal, and ileal mucosa. Slfn1 is a part of the Group I Slfn family and was increased in expression in both duodenal and jejunal mucosa of male and female VC-Slfn3KO mice compared to Ctrl mice ( Fig 4A). Slfn2, the other murine member of Group I, was increased in expression in female VC-Slfn3KO mice compared to Ctrl mice, but not in male mice (Fig 4B). Neither Slfn1 or 2 had expression differences between VC-Slfn3KO and Ctrl mice in the ileal mucosa ( Fig  4B and 4C). Slfn4, a Group II Slfn closely related to Slfn3, was increased in expression in the duodenal and ileal mucosa of both male and female VC-Slfn3KO but did not change in expression in jejunal mucosa in comparison to the Ctrl mice ( Fig 4C). Finally, the Group III Slfn members Slfn5 and 9 displayed a decrease in expression in both male and female VC-Slfn3KO mice in the ileal region and Slfn8 only increased significantly in female VC-Slfn3KO jejunum (Fig 4D-4F).
We also evaluated the expression of Slfn family members in the spleen. We did not expect to see a change in Slfn family expression in these VC-Slfn3KO mice due to the Vil-Cre being epithelial-specific for knocking out Slfn3, as we had seen in the pan-Slfn3KO mice [18]. Slfn3 RNA expression was present in all groups (S3 Fig). There was an overall trending decrease in Slfn family expression in the VC-Slfn3KO mice and a significant decrease in Slfn9 in the female VC-Slfn3KO mice in comparison to Ctrl mice (S3 Fig). Slfn4 RNA expression was undetected in both Ctrl and VC-Slfn3KO splenic tissues.

Differentiation marker expression of SI, Vil1, and Dpp4 are influenced by epithelial Slfn3 loss and by sex
We have chosen to study SI, Vil1, and Dpp4 as markers of intestinal differentiation. SI and Dpp4 are brush border digestive enzymes and have been established as canonical markers for enterocytic differentiation [22][23][24][25]. Vil1 is found within the microvillus core of the brush border and is a key calcium regulated actin binding protein [26][27][28][29]. In the duodenum, female VC-Slfn3KO mice exhibited increased SI, Vil1, and Dpp4 expression compared to controls (Fig 5A-5C), but male knockouts did not. A similar female-only increase in differentiation markers was observed in the jejunum of the female VC-Slfn3KO except that SI was similarly expressed between male and female Ctrl and VC-Slfn3KO mice. SI was increased in the ileal mucosa of both male and female VC-Slfn3KO mice but surprisingly Vil1 was decreased in the female VC-Slfn3KO mice while Dpp4 did not differ significantly among the groups. Since SI had such a marked change in RNA expression in both male and female VC-Slfn3KO mice, we evaluated SI protein expression by flow cytometry in EpCam + epithelial cells, excluding CD3 + immune cells (which also express Slfn3 and SI). SI protein expression changed based on sex and intestinal region. Enterocytes from male VC-Slfn3KO mice did not achieve a statistically

PLOS ONE
Vil-Cre specific Schalfen 3 knockout mice exhibit sex-specific differences in intestinal genes significant change in SI protein levels compared to Ctrl mice. However, female VC-Slfn3KO mice had significantly less SI protein in jejunal enterocytes compared to Ctrl mice (Fig 5D and  5E). The CD3 + SI + immune cells were evaluated as well, and we found that there was a decrease in SI protein in the duodenum of female VC-Slfn3KO mice (S4 Fig).

PLOS ONE
Vil-Cre specific Schalfen 3 knockout mice exhibit sex-specific differences in intestinal genes

Glucose transporter expressions were impacted by epithelial Slfn3 loss and by sex
Glucose transporters are also important membrane proteins that serve as markers of intestinal epithelial differentiation and functional maturation [11,30,31]. Female VC-Slfn3KO mice exhibited increased SGLT1, Glut1, and Glut2 expression in the duodenum and increased

PLOS ONE
Vil-Cre specific Schalfen 3 knockout mice exhibit sex-specific differences in intestinal genes SGLT1 and Glut2 in the jejunum, while male mice did not (Fig 6). In the ileum, we observed decreased SGLT1 expression in both male and female VC-Slfn3KO mice, but increased Glut1. Glut2 did not change significantly (Fig 6).

Discussion
This study investigates the specific role of Slfn3 in intestinal cell differentiation. By creating an epithelial-specific knockout of Slfn3, we were able to investigate the effects of Slfn3 loss in intestinal epithelial cells without possible effects of the loss of Slfn3 in other cell types that could have been observed in the pan-Slfn3KO mouse [16]. We observed that the loss of Slfn3 affected the RNA expression levels of Slfn family members and intestinal differentiation markers based on small intestinal region and sex.
Previously, in pan-Slfn3KO mice, we have shown that the villus length was decreased in male Slfn3KO mice and increased in female Slfn3KO mice in comparison to wild type mice. The crypt depth decreased in both male and female Slfn3KO mice [32]. We observed these same changes to the villus length and crypt depth with the VC-Slfn3KO mice. Additionally, when we transiently interfered with Slfn3 expression in the intestine of male rats by injecting specific siRNA into the lumen of temporarily blocked jejunal segments, villus length was decreased, but there was no change in crypt depth [12]. We and others have published Slfn3 exhibiting antiproliferative properties and that when Slfn3 is lost by siRNA knockdown or increased by ectopic expression or adenovirus, cell proliferation directly correlates [7,15,33].

PLOS ONE
Therefore, these in vivo data further indicate a sex-dependent role for Slfn3 in the regulation of villus length and crypt depth.
Previous evaluation of Slfn family member RNA expression in the pan-Slfn3KO mice displayed an increase in Slfn1 and Slfn5 and a decrease in Slfn4, Slfn8, and Slfn9 in the ileal mucosa and was not affected by sex [18]. In this study, we extended our assessment to all intestinal regions and found a sex-dependent increase for Slfn2 and Slfn8 in the duodenal and jejunal regions of the female VC-Slfn3KO vs. Ctrl mice. Ileal Slfn9 decreased in expression for VC-Slfn3KO similarly to what was seen in pan-Slfn3KO. Conversely, ileal Slfn4 expression increased and ileal Slfn5 expression decreased in VC-Slfn3KO, whereas they were oppositely expressed in the pan-Slfn3KO ileum [18] (Table 2). Other studies have shown that Slfn family expression can be induced or reduced due to IFN treatment. Slfn1, 2, 5, and 8 mRNA expression is induced after treatment with IFNα and that knockout mouse embryonic fibroblast cell lines for Stat1 is required for all IFN-inducible Slfn genes, Stat3 is required for all Slfns expect Slfn5, and p38 MAPK is required for Slfn1 and 2 but not Slfn5 and 8 [7,34]. Additionally, human SLFN5 and SLFN11 are induced by IFN-β, IFN-γ, poly-inosine-cytosine or poly dAdT [35,36]. We evaluated Slfn family expression in spleen tissue from the VC-Slfn3KO mice. Slfn3 expression was detected in the splenic tissue, which was to be expected in this epithelialspecific KO, however, we did not expect to see a trending decrease in the Slfn family members and a significant decrease in Slfn9 of the VC-Slfn3KO females. A possible explanation could be the fact that Slfn paralogues are clustered together and have close sequence similarity, which can lead to potential redundancy in their expressions, functions, and unclear identification of which paralogue is responsible for a particular phenotype [1,2,37].

PLOS ONE
Vil-Cre specific Schalfen 3 knockout mice exhibit sex-specific differences in intestinal genes that were injected intraluminally with either adenoviral Slfn3 or Slfn3 siRNA. In these studies, the RNA and protein expression of SI, Vil1, and Glut2 increased with the exposure to Slfn3 adenovirus while suppression of Slfn3 by siRNA led to a decrease in these same markers [12].
In contrast, we found that in the VC-Slfn3KO mice SI RNA expression was increased in the ileum and in the duodenum only for female VC-Slfn3KO mice. But then after protein evaluation, we only found a decrease in SI protein in the jejunum of female VC-Slfn3KO mice. This contrasting SI RNA and protein expression could be similar to how we have previously shown that the human homolog SLFN12, regulates differentiation markers in Caco2 cells and nonmalignant HIEC-6 enterocytes via a pathway involving SerpinB12, deubiquitylases UCHL5 and USP14, and the transcription factor Cdx2. Furthermore, Slfn12 regulates ZEB1 and Slug protein conversely to the mRNA in breast cancer cells and c-Myc in lung adenocarcinoma cells [39,40]. One other possibility could also be that the SI RNA expression was measured from mucosa which includes epithelial and immune cells. Whereas the SI protein expression was measured in only EpCam + epithelial cells and the CD3 + immune cells were gated out. However, the protein expression of SI in the CD3 + immune cells does not mirror the overall mucosal SI protein levels. Therefore, the differences seen between SI RNA and protein expression could be the combination of expressions between mucosal versus epithelial cells or it could indicate how Slfn3 regulates SI expression at the translational level possibly similarly to SLFN12.
The influence of gender on the expression of intestinal genes and differentiation markers has been less studied in mice and humans but has been most extensively studied in Drosophila. A human metabolic syndrome study found that women had decreased plasma Dpp4 activity levels in the metabolic syndrome patients compared to control patients [41]. Renal Glut1 and SGLT1 mRNA expression levels have been shown to be higher in female mice compared to male mice [42]. Interesting work done in Drosophila midgut, by Hudry et al. revealed that there are sex differences in intestinal stem cell proliferation that was adult-reversible and intrinsic to the stem cells [43]. They went on to determine that in females, genes associated with cell division-related processes were more greatly expressed, while males had a greater expression of genes coding for proteins that function in carbohydrate metabolism and redox processes [43]. These initial studies suggested that sex-biased intestinal metabolism could contribute to the sex differences observed in whole-body physiology. A subsequent study revealed a bi-directional communication between the male gonad and the adjacent intestinal midgut [44]. This study showed that the testis promotes sex differences in carbohydrate metabolism via a type I family of cytokines called Unpaired (Upd) in Drosophila, which are similar to mammalian interleukins and leptin (which are also highly sexually dimorphic in humans and rodents) [44]. The testis-secreted cytokines promoted enterocyte JAK-STAT signaling within the adjacent intestinal midgut leading to sugar gene expression and gut-derived citrate which promoted food intake and sperm maturation [44]. Correlatively, the JAK-STAT pathway is the most important pathway in regulating IFN-inducible genes [34,[45][46][47][48]. Therefore, since Slfn expression is induced by IFNs [34], it could be possible that the sex-biased gene expression changes that we observed in this study could be due to an interruption of JAK-STAT signaling with the loss of Slfn3.

Conclusions
Overall, our targeted deletion of Slfn3 using a Vil-Cre mouse model allowed us to evaluate the effects of Slfn3 loss on Slfn family members and intestinal differentiation markers in intestinal epithelial cells. We not only found that the loss of Slfn3 did influence the expression of Slfn family members and intestinal differentiation markers, but also that this influence varied along the length of the small intestine and was largely sex-biased to females. While the focus of this study was on the small intestine, future investigations may consider how this epithelial Slfn3 knockout affects other epithelia such as gastric, colonic, renal, and pulmonary. Additional future studies could also explore sex-secreted cytokines possibly influenced by the JAK-STAT pathway. Further determining the pathways that Slfn3 is involved in will allow us to focus pathway analysis in the human ortholg, SLFN12. We could then target these pathways for treatment in intestinal diseases, including intestinal diseases that display gender bias, such as inflammatory bowel disease in women.