Hyperplasia of Interstitial Cells of Cajal in Sprouty Homolog 4 Deficient Mice

Gastrointestinal stromal tumors, which are thought to derive from interstitial cells of Cajal or their precursors, often harbor an oncogenic mutation of the KIT receptor tyrosine kinase. Sprouty homolog 4, a known negative regulator of ERK pathway, has been identified in the interstitial cells of Cajal in the KitK641E murine model of gastrointestinal stromal tumors. Sprouty homolog 4 was upregulated both at the mRNA and protein level in these cells, suggesting that Sprouty homolog 4 is downstream of oncogenic KIT activation and potentially engaged in the negative feedback loop of ERK activation in this model. Here, we used KitK641E heterozygous and Sprouty homolog 4 knock out animals to quantify interstitial cells of Cajal in situ, using quantitative immunofluorescence for the receptor tyrosine kinase Kit and for phosphodiesterase 3a (PDE3A). In the antrum of Sprouty homolog 4 knock out mice, hyperplasia of interstitial cells of Cajal was reminiscent of the KitK641E heterozygous mice antrum. Additionally, the density of interstitial cells of Cajal was higher in the colon of adult Sprouty homolog 4 knock out mice than in WT littermates, although hyperplasia seemed more severe in KitK641E heterozygous mice. Functional transit studies also show similarities between Sprouty homolog 4 knock out and KitK641E heterozygous mice, as the total transit time in 9 month old animals was significantly increased in both genotypes compared to WT littermates. We concluded that the lack of Sprouty homolog 4 expression leads to hyperplasia of the interstitial cells of Cajal and is functionally associated with a delayed transit time.


Introduction
Gastrointestinal stromal tumors (GIST) are the most common sarcoma of the gastrointestinal tract. They are thought to derive from the interstitial cells of Cajal (ICC), or an ICC precursor, and are highly resistant to conventional chemotherapy and radiotherapy [1,2]. Approximately

Ethics statement
Study of Kit K641E and Spry4 KO mice was approved by the ethics committee for animal wellbeing of the Faculty of Medicine, Université Libre de Bruxelles (Protocol number 491N).

Animals
Generation of Spry4 KO and Kit K641E animals have been described by Klein et al. 2006 [26] and Rubin et al. 2005 [22], respectively. In our colony, Kit K641E/K641E and Kit K641E/K641E -Spry4 KO mice died around 12 days after birth. Therefore, only WT, Kit WT/K641E and Spry4 KO genotypes could be investigated in adult (3 and 9 month old) mice. Genotyping was performed as described [22,26], with primers listed in Table 1. Body weight was determined before mice were sacrificed by cervical dislocation and decapitation. Small intestine, colon and stomach were promptly removed. The gastric antrum was delineated from corpus based on visual landmarks on the serosa. Luminal content was gently emptied and surrounding tissues (e.g. mesenteric fat) were carefully removed by sharp dissection without damaging the serosa.

Real time quantitative PCR (qPCR)
A minimum of three different RNA samples from P10 WT, Kit K641E/K641E and Spry4 KO mice antrum were used. Total RNA was extracted using RNeasy MiniKit (Qiagen, Valencia, CA, USA) according to manufacturer's instructions. Genomic DNA was removed using the RNase-Free DNase set (Qiagen). RNA was reverse transcribed with 200 units of M-MLV Reverse Transcriptase (Invitrogen, Eugene, Oregon, USA) in a reaction containing 1μg of random primers (Amersham Bioscience, Piscataway, NJ, USA), 10mM each dNTP, 1x First-Strand buffer and 100mM dithiothreitol followed by heat deactivation. The cDNA reverse transcription product was amplified with specific primers ( Table 2) by qPCR using SYBR Green chemistry on a 7500 Real-time PCR system (Applied Biosystems, Foster City, CA, USA). Identical thermal profile conditions, namely 95°C for 10min, then 40 cycles of 95°C for 15sec and 60°C for 1min were used for all primer sets. Emitted fluorescence was measured during annealing/ extension phase and amplification plots were generated using the Sequence Detection System. Transcriptional quantification relative to GAPDH and β-actin reference genes was performed using qBase+ software (Biogazelle, Zwijnaarde, Belgium).

Immunofluorescence (IF)
Tissues were fixed for 24h in fresh 4% paraformaldehyde, pH 7.4, and cryopreserved in sucrose solutions (10%, 20%, 30% w/v in water), overnight (o/n) each, embedded in OCT (Sakura Finetec Europe, Leiden, the Netherlands) and frozen at -80°C. Circumferential sections (16μm Table 1. Primers used for genotyping. IF was carried out as described [28]. Briefly, slides were brought to room temperature (RT), permeabilized and blocked for 1 hour in 10mM TBS pH 8.2 containing 0.1% Triton X-100 (Sigma, Saint Louis, MO, USA) and 10% normal horse serum (NHS). Primary antibodies were diluted in a TBS-Triton X-100 0.1% and 1% NHS solution and incubated overnight at RT in a humid chamber. Slides were washed in TBS and incubated at RT for 1 hour in TBS containing the secondary antibodies. Slides were washed and mounted using Glycergel (Dako, Glostrup, Denmark) + 2.5% DABCO (Sigma). SPRY4 antibody specificity was assessed by preabsorption with the antigenic peptide. In short, SPRY4 antibody was incubated in the presence of 0.1μg/ ml or 1μg/ml SPRY4 immunogenic peptide (Santa Cruz Biotechnology, Inc., Dallas, TX, USA, sc-18607P) for 30 min at 4°C. Subsequently, IF was carried out as described above.
Primary and secondary antibodies used for IF are summarized in Table 3. Slides were observed and imaged on an AxioImager Z1 fluorescent microscope (Zeiss, Jena, Germany), using a Plan Apochromat 20x/0.8 or EC Plan NeoFluar 40x/0.75 objective. Excitation was provided by a HBO 105W lamp. Band pass filters sets #49, #38; and #43 (Zeiss) were used to detect blue, green and red fluorochromes respectively. Images (1388 by 1040 pixels, pixel size (x-y): 0.32 micron by 0.32 micron) were acquired sequentially with an Axio-CamMRm camera (Zeiss) as 3 x 12 bit RBG proprietary Ã .zvi files. Files were processed with AxioVision (4.6) software (Zeiss). Images were displayed in linear mode with manual contrast adjustment and exported as uncompressed. TIF files. Figures were prepared with Adobe Illustrator.
Quantification of PDE3A-ir, KIT-ir and HuC/D-ir was performed using the Fiji software [29]. The plugin "Stich grid of images" [30] was used to assemble images covering the entire circumference of the sample. The boundaries of the muscularis propria were delineated to extract the region of interest (ROI). The area of immunoreactivity for the ICC or neural body markers, KIT-ir and PDE3A-ir or HuC/D-ir, respectively, was determined by thresholding within the ROI of the muscularis propria.

Confocal microscopy
High resolution confocal imaging was performed using a Zeiss LSM780 system fitted on an Observer Z1 inverted microscope equipped with a LD LCI C-Apochromat 40x/1.1 W objective (Zeiss). The 488 nm excitation wavelength of the Argon/2 laser, a main dichroic HFT 488 and a band-pass emission filter (BP500-550 nm) were used for selective detection of the green fluorochrome. The 543nm excitation wavelength of the HeNe1 laser, a main dichroic HFT 488/ 543/633 and a long-pass emission filter (BP565-605 nm) were used for selective detection of the red fluorochrome. A 405 nm blue diode, a main dichroic HFT 405 and a band-pass emission filter (BP435-485 nm) were used for selective detection of the DNA counterstain.
Z-stacks of images were acquired sequentially with a zoom factor of 2 and optimal (1 Airy unit) pinhole (scaling (x-y-z): 0.21 x 0.21 x 0.53 micron) and stored as 8-bit proprietary Ã .czi files. Single plane images were displayed using Zen2010 software (Zeiss) and exported as 8 bits uncompressed Ã .TIF images. Figures were prepared as above.

Total gastrointestinal transit time
Total gastrointestinal transit time was carried out as described [31]. Briefly, 200μl of 6% (w/v) carmine red (Sigma) suspended in 0.5% methylcellulose (Sigma) was administered by gavage through a 20 gauge round-tip feeding needle (n = 4-11 in each genotype). The time at which gavage took place was recorded as T 0 . After gavage, fecal pellets were monitored for the presence of carmine red at 10 min intervals. The interval between T 0 and the time of first observance of carmine red in stool was considered as total gastrointestinal transit time.

Statistics
All data represent the mean ± SD. Statistical analysis was performed with Prism 6 software (GraphPad Software, Inc., La Jolla, CA, USA), using Kruskal-Wallis test with Dunn's post hoc test to compare different columns. A p-value smaller than 0.05 was regarded as statistically significant.

SPRY4 immunostaining in antrum of Kit K641E/K641E and Kit WT/K641E mice
This study was based on our previous observation of elevated Spry4 expression in Kit K641E/K641E antrum as compared to WT mice [19]. Here, we confirmed, in 10-day-old animals (P10), strong Spry4 expression in Kit K641E/K641E antrum compared to WT littermates, while no Spry4 expression could be detected in Spry4 KO or Kit K641E/K641E -Spry4 KO antrum (S1A Fig). In 3-monthold animals, SPRY4-ir was detected in the ICC of Kit WT/K641E animals, but not in WT nor in Spry4 KO animals (Fig 1). These data are in line with our previous observations [19]. SPRY4 immunoreactivity (-ir) was also detected in the ICC of P10 Kit K641E/K641E mice, in contrast to WT and Spry4 KO mice (S1B Fig).
To validate the SPRY4 antibody used in this study (Table 3), preabsorption with an immunogenic peptide was performed on P10 Kit K641E/K641E antrum to assess the disappearance of SPRY4-ir signal after preabsorption (S2 Fig).

Significant increase of ICC area in antrum of 3-month-old Spry4 KO mice
Phosphodiesterase 3A (PDE3A) has previously been identified as an ICC marker [19,32]. Double immunofluorescence staining using KIT (rabbit) and PDE3A (sheep) antibodies (Table 3) confirmed the concordance of these markers in 3-month-old antrum (S4 Fig). PDE3A-ir filled in the cytoplasm while KIT-ir was to a large extend present at the cell membrane but the two clearly labelled the very same cells (S4 Fig). Since many antibodies used in this study (Table 3) have been raised in rabbit, the sheep PDE3A antibody was preferred as ICC marker for subsequent double immunofluorescence studies.
Quantification of the PDE3A-ir ICC density in the musculature was performed on the entire circumference of the antrum, by normalizing the area of PDE3A-ir positive signal to the total muscularis propria area. Compared to WT littermates, 3-month-old Spry4 KO animals showed a significant increase in PDE3A-ir ICC area, similar to Kit WT/K641E mice (Fig 2) while in P10 Spry4 KO animals, no difference was observed in the PDE3A-ir ICC area compared to their WT littermates (S5 Fig). In contrast, P10 homozygous Kit K641E/K641E already exhibited a massive PDE3A-ir ICC hyperplasia (S5 Fig), replacing the entire longitudinal muscle layer in the antrum. As these homozygous mice died before weaning, heterozygous Kit WT/K641E mice, which display a milder phenotype of ICC hyperplasia, were used for subsequent comparative studies in adult animals.

No change detected in pERK in antrum ICC of Spry4 KO mice at P10 and 3 months
With Sprouty proteins being known as negative regulators of the ERK pathway [14], we hypothesized that phosphorylation of ERK might be elevated in the hyperplastic ICC of the Spry4 KO animals compared to WT mice. We observed pERK-ir in PDE3A-ir ICC of P10 Kit K641E/K641E animals, but not in WT nor in Spry4 KO littermates (S6 Fig). In 3-month-old animals, pERK-ir was not detected in PDE3A-ir ICC of Spry4 KO or Kit WT/K641E antrum. Conversely, in all genotypes at any age, robust pERK-ir was consistently detected in the myenteric plexus and in intramuscular nerve fibers adjacent to PDE3A-ir ICC in the same field of view ( Fig 3) and was thus regarded as an internal positive control.

No compensation by other SPRYS for the loss of SPRY4 in antrum
We next wondered if compensation mechanisms for the loss of SPRY4 would be present. Therefore, we tested by qPCR the expression levels of SPRY1, SPRY2 and SPRED1 but none of these genes showed any upregulation in Spry4 KO antrum ( Fig 4A). Additionally, SPRY2-ir was detected only in the smooth muscle cells of antrum and not adjacent PDE3A-ir ICC ( Fig  4B, S7 Fig).

No change in other signaling pathways in ICC of 3-month-old antrum
A role of SPRY4 in PI3K/AKT/mTOR signaling has been established in other models [33,34]. pAKT-ir has previously been reported in the hyperplastic ICC layer in P10 Kit K641E/K641E antrum [35] and (S8 Fig). In 3 month old animals of all genotypes, pAKT-ir was detected only in the myenteric plexus but not in PDE3A-ir ICC (Fig 5).
To further investigate a possible involvement of mTOR pathway, we performed IF for the active form of p70S6, a protein downstream of mTOR, phospho-p70S6 (pp70S6). pp70S6-ir. but, in all other genotypes, at both P10 and 3 month old, pp70S6-ir was detected only in the myenteric plexus of the antrum but not in PDE3A-ir ICC (Fig 6).
Although SPRY4 has not been directly implicated in the JAK/STAT pathway, this pathway is also involved upon KIT receptor activation [35][36][37][38]. Immunoreactivity for the active (phosphorylated) form of STAT5 (pSTAT5-ir) was detected in the hyperplastic PDE3A-ir ICC layer   Hypoganglionosis in antrum of both Spry4 KO and Kit WT/K641E mice Hyperganglionosis has been reported in the colon of Spry2 KO mice [39]. We thus wondered whether the enteric nervous system (ENS) could similarly be affected in Spry4 KO mice. HuC/ D-ir was used for the detection of myenteric neural bodies [40] in the antrum of 3-month-old mice (Fig 8A). HuC/D-ir positive area was normalized for the muscularis propria area. Both Spry4 KO and Kit WT/K641E mice showed a decrease of neural bodies in antrum, compared to WT (Fig 8B).

No change in ICC or signaling pathways in small intestine of 3-month-old Spry4 KO mice
Hyperplasia of ICC in Kit K641E animals is not restricted to the antrum [22]. Hence, we also performed PDE3A immunostaining in order to quantify ICC in the small intestine (Fig 9, as described above). A trend towards an increase of PDE3A-ir ICC area in small intestine of Spry4 KO mice did not reach statistical significance (n = 5-7 animals per group) while, the PDE3A-ir ICC area was significantly increased in Kit WT/K641E small intestine (Fig 9).
In the small intestine, in all genotypes, pERK-ir was detected only in the myenteric plexus (S11 Fig), and SPRY2 solely in the smooth muscle layers (S12 Fig). Quantification of neuronal bodies using HuC/D IF in 3-month-old small intestine did not show significant difference for any genotype (S13 Fig). Significant increase of ICC area in colon of 3-month-old Spry4 KO mice Double immunofluorescence staining using KIT (rabbit) and PDE3A (sheep) antibodies (Table 3) indicated an increased ICC area in both Spry4 KO and Kit WT/K641E colon (Fig 10A). Quantitative assessment of ICC area was performed by single IF staining using KIT-ir (two different, rabbit and goat, KIT antibodies- Table 3) and PDE3A-ir on adjacent sections of colon for WT, Spry4 KO and Kit WT/K641E genotypes (n = 5-7 animals per group). Within each genotype, the 3 antibodies gave concordant results, with non-significant differences between antibodies. The 3 antibodies identified similarly a significant increase in ICC area in Spry4 KO colon (p value <0.05 for each antibody) and in Kit WT/K641E colon (KIT rabbit p value <0.001, KIT goat and PDE3A p value < 0.01) compared to WT. Although ICC hyperplasia appeared more pronounced in Kit WT/K641E than in Spry4 KO, differences between Kit WT/K641E and Spry4 KO were not significant (p value PDE3A = 0.2509, p value KIT rabbit = 0.2668, p value KIT goat = 0.3572). Similarly to antrum and small intestine, pERK-ir was only detected in the myenteric plexus (S14 Fig), and SPRY2-ir solely in the smooth muscle layers (S15 Fig). HuC/D-ir quantification in the colon did not show any difference between genotypes (S16 Fig). Significant increase in total transit time in Spry4 KO and Kit WT/K641E mice ICC regulate the peristaltic movement of the gastrointestinal tract [41]. We wondered if ICC hyperplasia in Spry4 KO and Kit WT/K641E mice, could impact the gut propulsive function. We thus tested the transit time using the carmine red method in 3 and 9 month old animals. At 3 month of age, total transit time was similar between genotypes ( Fig 11A). Conversely, at 9 months, total transit time was significantly increased in both Spry4 KO and Kit WT/K641E , compared to their WT littermates (Fig 11B). Changes in total transit time were not related to the length of the small intestine since it did not differ between genotypes (S17 Fig).

Discussion
Here, we report a hyperplasia of ICC in both antrum and colon of SPRY4 KO mice. ICC were labeled using KIT-ir, the reference ICC marker [42] & (Fig 1 and Fig 10, S1 Fig and S4 Fig), but also using PDE3A-ir. Pde3a belongs to the gene expression profile of KIT-ir ICC cell-sorted in the mouse small intestine [32]. Pde3a also appeared among the genes upregulated in Kit K641E/ K641E antrum, presenting massive hyperplasia of KIT-ir ICC, compared to WT littermate and PDE3A-ir localized in KIT-ir ICC in the antrum of WT and Kit K641E (homozygous and heterozygous) mice. PDE3A-ir was therefore regarded as a novel marker for the KIT-ir ICC in the mouse gut [19]. Although PDE3A-ir has not gained yet a wider acceptance as ICC marker, to the best of our knowledge, our original claim [19] remains so far unchallenged in the literature. In the present study, we used high resolution confocal imaging to show, at the level of individually identified cells, that PDE3A-ir decorates electively the KIT-ir ICC, in Spry4 KO antrum, as well as in WT and Kit WT/K641E antrum (S4 Fig). We also quantitated the ICC area using KIT-ir (2 different antibodies) and PDE3A-ir in SPRY4 KO colon, as well as in WT and Kit WT/K641E (Fig 10). Within each genotype, KIT-ir and PDE3A-ir values proved to be concordant and, both PDE3A-ir and KIT-ir similarly detected the well-established ICC hyperplasia in Kit WT/K641E colon and revealed a significant ICC hyperplasia in Spry4 KO colon compared to WT. Although rare events, i.e. the possible occurrence of tiny populations of KIT negative PDE3A-ir cells or KIT-ir PDE3A negative cells, cannot be totally ruled out by these experiments, the presence of PDE3A-ir selectively in the KIT-ir ICC in the 3 genotypes studied and the ability of PDE3A-ir to detect, to the same extend as KIT-ir, the predicted ICC hyperplasia in the reference model for KIT-ir ICC hyperplasia, Kit K641E mice, and in Spry4 KO colon comfort our view [19] that PDE3A-ir represents globally a valuable marker for the Kit-ir ICC in the WT mouse gut and in ICC hyperplasia models.
Sprouty proteins are known negative regulators of the ERK pathway [14], a pathway involved in cell division, survival and transformation [43], we anticipated that ERK would play a critical role in the development of ICC hyperplasia. ICC represent less than 1% of the total cell population of the highly heterogeneous gut wall. Modifications in gene or protein expression occurring in ICC might thus vanish in the ambient 'noise' when using techniques which require tissue homogenization, e.g. qPCR or Western blot. Therefore, immunoreactivity was used to study signaling pathways in situ with resolution of the different cell types. Surprisingly, pERK was undetectable in the ICC of Spry4 KO mice. In order to exclude compensatory upregulation of other Sprouty family members, we performed qPCR of the antrum for Spry1, Spry2, and Spred1. None of these genes were upregulated in Spry4 KO antrum. No commercially available antibodies for SPRY1 and SPRED1 were fond suitable for IF on our material while SPRY2-ir showed no change between the different genotypes. Nevertheless, despite these limitations, our data suggest that no compensation by other SPRYs appears to occur. Recently, several papers indicated that SPRY4 can play a role in other signaling pathways [33,34,44,45]. Hence, pAKT and pp70S6 IF were performed in order to investigate the PI3K/AKT/mTOR and pSTAT5a/b for the JAK/STAT pathways, but no detectable differences in immunoreactivity could be seen between the different genotypes.
Interestingly, the phenotype of Spry4 KO mice is reminiscent of the phenotype of heterozygous Kit WT/K641E mice (Table 4). Both show hyperplasia of the ICC in antrum with no detectable changes in signaling pathways tested in this study. This contrasts with the homozygous Kit K641E/K641E phenotype, which shows complete replacement of the longitudinal layer of the antrum by a hyperplastic ICC layer in which pERK-ir, pAKT-ir, pp70S6-ir and pSTAT5a/b-ir is detectable. Rubin et al demonstrated elevated tyrosine phosphorylation of the KIT receptor in Kit K641E/K641E mice [22]. Hence, strong KIT phosphorylation in homozygous Kit K641E/K641E mice may lead to a strong activation of the signaling pathways, while heterozygous Kit WT/K641E mice would have lower levels of KIT phosphorylation and therefore lesser activation of the downstream signaling pathways than Kit K641E/K641E animals. In GIST882 cells, a human GIST cell line carrying the same K-to-E substitution as the Kit K641E mice [10], the link between ERK phosphorylation and upregulation of SPRY4 has been established [25,46]. One can only speculate that a small increase in ERK phosphorylation in the ICC of heterozygous Kit WT/K641E mice might well fall below the detection threshold of immunohistochemistry which readily picks up the strong ERK phosphorylation in ICC in Kit K641E/K641E mice and in the ENS in all genotypes. Despite the fact that SPRY4-ir was not detectable in ICC in the postnatal WT gut [19], Fig 1  & S1 Fig, this study revealed ICC hyperplasia in Spry4 KO mice, raising the possibility that Spry4 might play a role during ICC embryonic development. Noteworthy, Spry4 expression, detected by in situ hybridization, has been reported in mouse embryo, in stomach at embryonic day 11.5 (E11.5) and in intestine at E12.5 and E14.5 [47], i.e. around the time where ICC differentiation starts [48][49][50].
In a similar perspective, Taketomi et al reported that SPRY2 deficient mice exhibit hyperganglionosis in the colon, although the exact mechanism remains unclear [39]. Since, SPRY2 and SPRY4 belong to the same protein family [12,14], the ENS in Spry4 KO mice was also investigated. Surprisingly, the antrum of Spry4 KO mice presented a hypoganglionosis, while other parts of the gastrointestinal tract appeared normoganglionic. The same feature was also seen in Kit WT/K641E antrum, adding to the similarities between Spry4 KO and Kit WT/K641E ( Table 4). Migration of the neural crest cells forming the ENS is completed by E15 in mice [51]. Hence, the ENS alterations observed in Spry4 null and Kit WT/K64E animals after birth must originate during embryonic gut development.
The present study focused on the postnatal gut phenotype of Spry4 KO mice and further studies are clearly needed to unravel the time windows at which Spry4 is expressed in the different cell types and the underlying signaling pathways in the developing gut.
ICC are the pacemaker cells of the gastrointestinal tract, coordinating the contractility of the gastrointestinal muscle layers [41,52]. Bellier et al reported that PRM/Alf mice, which exhibit a higher number of ICC and increased intestine length, have a total gastrointestinal transit time similar to their WT littermates, implying a faster transit [53]. In contrast to PRM/Alf mice, the length of intestine in Spry4 deficient and Kit WT/K641E mice was similar to WT littermates and the increase in ICC was associated with a significantly delayed total gastrointestinal transit in aging (9 month old)-but not in younger (3-month-old) animals. This provides an original clue that, besides roles during development, Spry4 and Kit K641E may play additional roles during the aging process. Digestive transit is a very complex, multifactorial, process, and further studies (e.g. electrophysiology, microbiome, etc.) will be required to unravel the precise mechanism underlying these observations. Noteworthy, constipation, albeit a fairly common and unspecific complain, is a frequent symptom in human adult GIST patients [54][55][56][57].
In summary, we have shown that SPRY4 loss of function was associated with ICC hyperplasia in antrum and colon. The Spry4 KO mice bear striking similarities with the Kit WT/K641E oncogenic mice, and in both models, ICC hyperplasia was associated with a delayed total gastrointestinal transit time in aging mice.