Endocrinology and Metabolism

Colocalization of ghrelin O-acyltransferase and ghrelin in gastric mucosal cells

Ichiro Sakata, Jing Yang, Charlotte E. Lee, Sherri Osborne-Lawrence, Sherry A. Rovinsky, Joel K. Elmquist, Jeffrey M. Zigman


Ghrelin is a peptide hormone with many known functions, including orexigenic, blood glucose-regulatory, and antidepressant actions, among others. Mature ghrelin is unique in that it is the only known naturally occurring peptide to be posttranslationally modified by O-acylation with octanoate. This acylation is required for many of ghrelin's actions, including its effects on promoting increases in food intake and body weight. GOAT (ghrelin O-acyltransferase), one of 16 members of the MBOAT family of membrane-bound O-acyltransferases, has recently been identified as the enzyme responsible for catalyzing the addition of the octanoyl group to ghrelin. Although the initial reports of GOAT have localized its encoding mRNA to tissues known to contain ghrelin, it is as yet unclear whether the octanoylation occurs within ghrelin-producing cells or in neighboring cells. Here, we have performed dual-label histochemical analysis on mouse stomach sections and quantitative PCR on mRNAs from highly enriched pools of mouse gastric ghrelin cells to demonstrate a high degree of GOAT mRNA expression within ghrelin-producing cells of the gastric oxyntic mucosa. We also demonstrate that GOAT is the only member of the MBOAT family whose expression is highly enriched within gastric ghrelin cells and whose whole body distribution mirrors that of ghrelin.

  • membrane-bound O-acyltransferases
  • octanoate

the peptide hormone ghrelin was first described in 1999 following its isolation from stomach extracts (11). Functions that have been ascribed to ghrelin are many and include mediation of body weight and blood glucose homeostasis, mood and anxiety, growth hormone release, and memory, among others (3, 11, 16, 23, 30, 32). Ghrelin's ability to function in those capacities is dependent on the presence of the growth hormone secretogogue receptor (GHSR), which is the only ghrelin receptor to have been definitively identified, and also presumably binding of ghrelin to GHSR (3, 16, 24, 32). Ghrelin's interaction with GHSR requires a unique posttranslational modification to ghrelin involving the enzymatic addition of an O-octanoyl group to residue Ser3 (12, 17). Furthermore, this O-acyl posttranslational modification is required for many of ghrelin's known actions. For instance, in a recent study in which vaccines targeting ghrelin were developed as a potential strategy for treatment of obesity, only those vaccines that neutralized the octanoylated form of ghrelin, and not those targeting des-acyl ghrelin, resulted in decreased body weight gain, decreased adiposity, and decreased feed efficiency (30, 33).

The enzyme that catalyzes this reaction was recently identified and has been named GOAT, for ghrelin O-acyltransferase (6, 29). The discovery of GOAT involved its identification as a member of a large family of membrane-bound O-acyltransferases (MBOATs), followed by expression cloning studies in which INS-1 insulinoma cells cotransfected with cDNAs encoding preproghrelin and MBOAT4 (later renamed GOAT), but not those cotransfected with cDNAs encoding preproghrelin and any of the other 15 MBOAT family members, resulted in the production of octanoylated ghrelin (29). In a separate study, treatment of a human medullary thyroid carcinoma (TT) cell line with small interfering (si)RNAs that targeted MBOAT4, but not siRNAs targeting other MBOAT family members, resulted in a decrease in ghrelin octanoylation (6). Also, octanoylated ghrelin was found to be undetectable in GOAT gene knockout mice (6). Another key piece of evidence suggesting that GOAT functioned as ghrelin's acylating enzyme was its identification within tissues known to express ghrelin. These included stomach, intestine, colon, and testis in the mouse and mainly stomach and pancreas (as well as several other tissues at much lower levels) in human (6, 29). Of note, the levels of GOAT mRNA in the stomach were found to be greater than two orders of magnitude less than those of ghrelin mRNA. Also of note, the initial descriptions of ghrelin did not identify the specific cells within those ghrelin-expressing tissues that contain GOAT. The goals of the present paper were to determine whether GOAT and ghrelin are indeed coexpressed within the same cells and also to provide a thorough analysis of the tissue distribution of all the members of the MBOAT family.



Animals were housed under 12:12-h light-dark per day in a temperature-controlled environment. They were fed standard chow diet (8664 F6 Rodent Diet; Harlan Teklad) and had free access to water. All animals and procedures used were in accordance with the guidelines and approval of the UTSW Medical Center Institutional Animal Care and Use Committees.

Dual-label in situ hybridization histochemistry and immunohistochemistry.

Mice were deeply anesthetized with an intraperitoneal injection of chloral hydrate (500 mg/kg) and subsequently perfused transcardially with diethylpyrocarbonate (DEPC)-treated 0.9% saline followed by 10% neutral buffered formalin. Adult stomachs and duodenums were removed, stored in the same fixative for 4–6 h at 4°C, immersed in 20% sucrose in DEPC-treated phosphate-buffered saline (PBS), pH 7.0, at 4°C overnight, and embedded in Tissue-Tek OCT compound (Sakura Finetechnical, Tokyo, Japan). Cryosections of the stomachs and duodenums were prepared at a 10-μm thickness using a cryostat and mounted onto SuperFrost slides (Fisher Scientific, Pittsburgh, PA). All slides were air-dried and stored in desiccated boxes at −20°C until use.

In situ hybridization histochemistry was performed as reported previously (31). Prior to hybridization, sections were fixed in 4% formaldehyde in DEPC-treated PBS, pH 7.0, for 20 min at 4°C, dehydrated in increasing concentrations of ethanol, cleared in xylenes for 15 min, rehydrated in decreasing concentrations of ethanol, and placed in prewarmed sodium citrate buffer (95–100°C, pH 6.0). While in the sodium citrate buffer, slides were placed in a commercial microwave oven for 10 min at 20–70% power. Afterward, they were dehydrated, as before, in graded ethanols and air-dried. A 35S-labeled cRNA probe specific for mouse GOAT was prepared, as below, and was diluted to 106 cpm/ml in a hybridization solution containing 50% formamide, 10 mM Tris·HCl, pH7.5, 5.0 mg tRNA (Invitrogen, Carlsbad, CA), 100 mM dithiothreitol, 10% dextran sulfate, 0.6 M NaCl, 0.5 mM EDTA, pH 8.0, and 1× Denhardt's solution. The slides with tissue sections had hybridization solution and coverslips applied and then were placed in a 57°C incubator for 12–16 h. Next, coverslips were removed, and the sections were washed with 2× SSC buffer and incubated in 0.002% RNase A (Roche Molecular Biochemicals, Indianapolis, IN) with 0.5 M NaCl, 40 mM Tris·HCl, pH 8.0, and 0.1 mM EDTA for 30 min, followed by a 30-min incubation in the same buffer minus RNase. The sections were then submitted to stringency washes as follows: 2× SSC at 50°C for 1 h; 0.2× SSC at 55°C for 1 h; 0.2 SSC at 60°C for 1 h. Afterwards, the sections were dehydrated in increasing concentrations of ethanol containing 0.3 M NH4OAc followed by a final submersion in 100% ethanol.

Immediately following this procedure, the slides were processed for ghrelin immunoreactivity. This involved washing the slides with tissue sections in PBS three times, pretreating them with 0.3% hydrogen peroxide in PBS for 30 min at room temperature, and then incubating them in 3% normal donkey serum (Equitech-Bio, Kerrville, TX) with 0.25% Triton X-100 in PBS (PBT-azide) for 1 h. Next, the slides were incubated overnight at room temperature in anti-ghrelin antiserum (Phoenix Pharmaceuticals, Belmont,CA; Code H-031-31, Lot no. R451-2; 1:10,000 in PBT). After washing in PBS, sections were incubated in biotinylated donkey anti-rabbit IgG (Jackson ImmunoResearch Laboratories, West Grove, PA; 1:1,000) for 1 h at room temperature, followed by incubation for 1 h in a solution of avidin-biotin complex (Vectastain Elite ABC Kit; Vector Laboratories, Burlingame, CA; 1:500) diluted in PBS. The sections were next washed in PBS and incubated in a solution of 0.04% diaminobenzidine tetrahydrochloride (DAB; Sigma, St. Louis, MO) and 0.01% hydrogen peroxide in PBS. After sections were washed in H2O and dried, the slides were placed in X-ray film cassettes with BMR-2 film (Kodak, Rochester, NY) for 3 days. Slides were then dipped in NTB autoradiographic emulsion (Kodak), dried, and stored in desiccated, foil-wrapped boxes at 4°C for 6 wk. Finally, slides were developed with Dektol developer (Kodak), dehydrated in graded ethanols, cleared in xylenes, and coverslipped with Permaslip (Alban Scientific, St. Louis, MO).

Control experiments to test the specificity of the immunohistochemistry reactions included preadsorption of the anti-ghrelin antiserum with 50 μg/ml murine acylated ghrelin (catalog no. C-et-004; Global Peptide, Fort Collins, CO; suspended in saline), and also separate reactions in which anti-ghrelin antiserum was omitted. Staining was not observed in these preadsorption (Supplemental Fig. S1B) or primary antibody omission controls (see online version of this paper for supplemental materials).

Generation of GOAT cRNA probes.

Two different mouse GOAT-specific cRNA probes were generated by RT-PCR amplification of mouse stomach total RNA with the following oligonucleotides primer pairs: probe A: primer 1, 5′-GACTTCCCTTTTACAAGGGC-3′ and primer 2, 5′-ACCCAGGTAGTATTCGGTGTAGTG-3′; probe B: primer 3, 5′-GGGAGGCTCCCTGTGTTCC-3′ and primer 4, 5′-GCTTCGGTTCCACTGCCTGG-3′. The amplified 391-bp (probe A) and 400-bp (probe B) PCR products, which encompass the 1- to 391-bp region and the 592- to 991-bp region of GOAT mRNA (acc. no. EU721729), respectively, were gel purified and then subcloned into pDP18 (CU Minus) Transcription Vector (Ambion, Austin, TX) according to the manufacturer's protocol. The sequence and directionality of the insert were confirmed by DNA sequencing at the core DNA Sequencing Facility at UTSW Medical Center. To generate antisense and control sense 35S-labeled cRNA to use as probes, the plasmids were linearized by restriction digestion and then subjected to in vitro transcription with either T3 or T7 RNA polymerases according to the manufacturer's protocol (Ambion). Preliminary single-label in situ hybrization histochemistry analyses in which gastric mucosa was tested separately with each of the two probes were performed and demonstrated that probe A provided a stronger signal with less background than probe B. Thus, probe A was used for the dual-label studies. Use of the control sense GOAT probes did not result in any detectable deposition of silver granules in a pattern resembling that of single cells either when used alone for single-label in situ hybrization histochemistry experiments (Supplemental Fig. S2) or when used in dual-label studies along with DAB visualization of ghrelin immunoreactivity (Fig. S1C).

Isolation of enriched pools of gastric ghrelin cells.

For these studies, we used stomach mucosal cells generated from our recently reported ghrelin-GFP (green fluorescent protein) reporter mice (line hrGFP10) that harbor a transgene constructed by inserting the coding sequence for humanized Renilla reniformis GFP into a bacterial artificial chromosome that spanned the entire coding region of mouse ghrelin and contained nearly 59.36 kb sequence upstream of the start codon and ∼103.65 kb downstream of the stop codon (20). Stomach mucosal cells were released from the mucosa by using previously described enzymatic and mechanical dissociation techniques (21). Afterward, these dissociated cells were passed through a 100-μm filter and collected by centrifugation at 1,500 rpm for 5 min. The cellular pellet was further digested in 0.25% trypsin-EDTA (Invitrogen) at 37°C for 5 min. After a washing in PBS, cells were resuspended in FACS buffer (3% fetal bovine serum, 0.5 mM EDTA, 0.1% BSA, 10 U/ml DNaseI, 20 mg/ml glucose in Hank's buffered salt solution). Mucosal cells from male hrGFP10 mice (8–9 wk of age) were isolated, pooled, and then sorted with a cell sorter (MoFlo; Dakocytomation USA) at the UTSW Medical Center Flow Cytometry Multi-user Core Facility. Cells were sorted based on size, complexity, and intensity of GFP fluorescence (at 530 nm) and fluorescence at 585 nm. Four independent FACS preparations were included in the subsequent analysis (4–6 mice were used for each independent FACS preparation). For each preparation, those cells with the most intense fluorescence were collected as the hrGFP-enriched pool, whereas those with the least intense fluorescence were collected as the hrGFP-negative pool. Depending on the preparation, these hrGFP-enriched and hrGFP-negative pools corresponded to slightly different percentages of sorted, living cells as follows. The hrGFP-enriched pool corresponded to 0.53, 0.92, 0.44, and 0.61% of sorted, living cells in preparations 1, 2, 3, and 4, respectively; the hrGFP-negative pool corresponded to 92.91, 91.48, 93.25, and 90.56% of sorted, living cells in preparations 1, 2, 3 and 4, respectively.

RNA extraction and quantitative PCR.

After FACS sorting, we adjusted the hrGFP-enriched pools and the hrGFP-negative pools from each preparation to contain a matched number of cells (specifically, we used 6,000, 12,000, 10,000, and 4,000 cells for preparations 1, 2, 3, and 4, respectively). The cells in each pool were collected by centrifugation at 4°C at 3,000 rpm for 10 min. Total RNA was prepared using an RNAqueous-Micro kit (Ambion) and treated with DNaseI (TURBO DNA-free, Ambion). First-strand cDNA was synthesized from DNaseI-treated total RNA using random hexamer primers in a 100-μl reaction using a TaqMan Reverse Transcription kit (Applied Biosystems, Foster City, CA). Specific primers for each gene were designed using Primer Express software (PE Biosystems, shown in Supplemental Table S1). Of note, the primers used to amplify preproghrelin mRNA were designed to span the insertion site of hrGFP within the ghrelin open-reading frame, thus excluding the possibility of inadvertent amplification of any hybrid hrGFP-ghrelin mRNA species that theoretically could be produced from the transgene. Real-time RT-PCR reactions were set up in triplicate for each gene in a final volume of 20 μl containing 4 μl of first-strand cDNA, 167 nM of the forward and reverse primers, and 10 μl of 2× SYBR Green PCR Master Mix (PE Biosystems). PCR reactions were carried out in 384-well plates using the ABI PRISM 7900HT Sequence Detection System (PE Biosystems).

We also performed similar quantitative PCR analyses on hrGFP-positive and hrGFP-negative pools derived from two different preparations of gastric mucosal cells of a second ghrelin-hrGFP reporter line (R4). The R4 line was made using a bacterial artificial chromosome that contained nearly 192.04 kb of sequence upstream of the ghrelin start codon and nearly the entire coding sequence of ghrelin, ending ∼1.17 kb upstream of the ghrelin stop codon (20). The first preparation consisted of mucosal cells from each of 12 female R4 mice (8–9 wk of age), and the second preparation was made from mucosal cells from each of 8 female R4 mice (8–9 wk of age). Results similar to those for the hrGFP10 preparations were observed (data not shown).

PCR analysis of tissue distribution of MBOAT family mRNAs.

Six-month-old male C57BL/6J mice were anesthetized and euthanized at the end of the dark phase. Various tissues were collected and snap-frozen in liquid nitrogen. Prior to freezing, the stomach, small intestine, and colon were flushed with cold PBS, after which the small intestine was divided into three equal lengths for duodenum (proximal), jejunum (medial), and ileum (distal). Each segment of the gastrointestinal tract was cut open, and the mucosa was carefully scraped off. Total RNA was prepared from pooled mouse tissues by using the RNA STAT-60 kit from Tel-Test (Friendswood, TX). Total RNA was treated with DNaseI and analyzed for mRNA expression of the indicated genes by using the TITANIUM One-Step RT-PCR Kit (Clontech, Mountain View, CA). Each reaction contained 1 μg of DNaseI-treated total RNA from different mouse tissues and primers shown in Supplemental Table S2. Also, negative control reactions lacking RNA were set up for each pair of primers. The cycling parameters were set as 94°C × 30 s, 60°C × 30 s, and 68°C × 30 s. The number of cycles was 30 (preproghrelin, ACAT1, ACAT2, GUP1, LRC4, MBOAT1, MBOAT2, and MBOAT5) or 35 (GOAT, DGAT1, HHAT, and PORC). Aliquots (20 μl) of the 50-μl RT-PCR samples were then loaded onto 1.5% agarose gel, and PCR-amplified bands were visualized by use of ethidium bromide.

Analysis of data.

Tissue sections were viewed with a Zeiss Axioskop 2 microscope under both darkfield and brightfield conditions, and photomicrographs were produced with a Zeiss digital camera attached to the microscope. To determine the degree of overlap between cells with GOAT mRNA expression and ghrelin immunoreactivity, three different sections of gastric fundus, each separated by ≥100 μm along the length of the fundus, were examined in each of four separate animals. Mean percent ± SE of ghrelin-immunoreactive cells coexpressing GOAT was determined by averaging the results obtained for each of the three sections of gastric fundus in the four mice mentioned above. A similar analysis was done with duodenum. To determine the density of ghrelin-immunoreactive cells per area of gastric or duodenal mucosa, we counted all the ghrelin-immunoreactive cells located within each of the three sections of stomach and duodenum described above for each of the four mice. The total number of ghrelin-immunoreactive cells for each section was then divided by the calculated total area of each mucosal section, as determined using the Scion Image for Windows software (Scion, Frederick, MD). Mean percent ± SE of ghrelin-immunoreactive cells per square millimeter was determined by averaging the results obtained for each of the three sections of gastric fundus in the four mice mentioned above as well as for each of the three sections of duodenal mucosa in the four mice mentioned above. Criteria used to determine whether a ghrelin-immunoreactive cell coexpressed GOAT mRNA included both 1) brightfield visualization of silver granules overlying the DAB-stained cell at ≥ 2×, ≥3×, and ≥5× the background density of silver granule deposition and 2) conformation of the overlying silver granules to the shape of the underlying DAB-stained cell. Criteria used to localize potential GOAT mRNA-containing cells lacking ghrelin immunoreactivity involved examining the same dual-labeled sections for areas that 1) contained ≥5× the background density of silver granule deposition and 2) conformed to the average size of a DAB-stained cell. The imaging/editing software program Adobe PhotoShop 7.0 (San Jose, CA) was used to adjust contrast, brightness, and color of the photomicropraphs.

The relative expression of mRNAs, as determined by quantitative PCR, was calculated using the comparative threshold cycle (CT) method. In particular, the mean expression levels for a particular mRNA species, as normalized to the housekeeping gene cyclophilin, was calculated using the following equation: 2(−x), whereby x was calculated by subtracting the CT value for the housekeeping gene cyclophilin from the CT value for the mRNA of interest (ΔCT). Such was done separately for each of the four hrGFP-positive pools and each of the four hrGFP-negative pools described above. We determined the means ± SE for each of the two pool types and then used Student's t-tests to determine statistical significance. P < 0.05 was considered statistically significant. In addition, we also compared the CT values of the genes of interest to the CT values of a separate housekeeping gene, 36B4, and observed similar results to those determined with cyclophilin (data not shown).


Histochemical evidence for localization of GOAT within gastric ghrelin cells.

To determine whether GOAT expression within the stomach colocalizes with that of ghrelin, we performed dual-label histochemical analyses on thin cryosections of mouse stomach. We visualized binding of GOAT mRNA-specific riboprobe together with ghrelin immunoreactivity on the same sections of gastric oxyntic mucosa. Our results indicate a high degree of colocalization of GOAT and ghrelin (Fig. 1, A–E; Table 1). Using usual parameters to determine expression of mRNA within a cell (overlying silver granules conform to the shape of the cell and appear in a density at ≥2× that of the background), we found that 71.2 ± 1.4% of ghrelin-immunoreactive gastric cells also expressed GOAT mRNA (Fig. 1; Table 1). Lower percentages were observed when overlying silver granule depositions at densities ≥3× or ≥5× that of the background were used to define colocalization. As such, there appeared to be a subset of ghrelin-immunoreactive cells that did not contain GOAT mRNA (Fig. 1, C and F). Also, there were occasional areas equivalent in size and shape to the DAB-stained ghrelin cells that appeared to have a concentrated number of overlying silver granules (indicating GOAT mRNA expression) but no DAB signal (indicating lack of ghrelin immunoreacitivity). Although difficult to determine without a counterstain such as DAB to identify the borders of the cells, these areas could represent cells that contain GOAT but lack ghrelin (Fig. 1, C and G). Using this method, we estimate that 14.5 ± 1.2% of all GOAT mRNA-containing gastric mucosal cells lacked coexpression of ghrelin immunoreactivity.

Fig. 1.

A series of photomicrographs demonstrating coexpression of GOAT (ghrelin O-acyltransferase) with ghrelin in the mouse gastric oxyntic mucosa. Ghrelin-immunoreactive cells are stained orange brown. Cells expressing GOAT mRNA have overlying silver granules, which alternatively appear black when viewed under bright-field methods (A, C, E, F, and G) or white when viewed using dark-field optics (B and D). Low-power (A and B) and high-power (C and D) photomicrographs of labeled cells within representative gastric section. The majority of the cells are dual labeled for both ghrelin immunoreactivity and GOAT mRNA, as evidenced by the representative magnified cell in E. On this particular section of gastric mucosa, there are also 5 cells that apparently contain only ghrelin immunoreactivity [surrounded by circles and as evidenced by the representative magnified cell in F and one presumed cell that contains only GOAT mRNA (G)]. Scale bars, 50 μm in A–D, 5 μm in E–G.

View this table:
Table 1.

Estimates of coexpression of GOAT mRNA in ghrelin-immunoreactive gastric cells

Histochemical evidence for localization of GOAT within duodenal ghrelin cells.

Another known site of relatively abundant ghrelin production is the duodenum. To determine whether GOAT also colocalizes with ghrelin within the duodenum, we repeated the dual-label histochemical analyses described above on duodenal mucosa. In our hands, we found that the density of ghrelin-immunoreactive cells within the duodenum was lower than that within the gastric fundus (17.0 ± 2.6 vs. 56.1 ± 2.4 cells/mm2). We also experienced a diminution in the numbers of detectable ghrelin-immunoreactive cells within the duodenum upon combining the methods for detection of ghrelin immunoreactivity with those for detection of GOAT mRNA (3.2 ± 0.4 cells/mm2) (Fig. 2); this phenomenon was not observed upon dual-labeling analyses of gastric mucosa (56.5 ± 6.4 ghrelin-immunoreactive cells/mm2). Nonetheless, we still found several ghrelin-immunoreactive cells that coexpressed GOAT mRNA (Fig. 2, B and 2B′). In particular, 19.4 ± 0.6% of ghrelin-immunoreactive duodenal cells contained overlying silver granules (representing binding of radiolabeled GOAT cRNA probe) at ≥2× the background density of silver granule deposition.

Fig. 2.

Histochemical demonstration of coexpressed GOAT and ghrelin within duodenal mucosal cells. A: ghrelin immunoreactivity is observed scattered throughout the duodenum in this representative section. B: although performing both in situ hybridization histochemistry for GOAT mRNA and immunohistochemistry for ghrelin on a nearby duodenal section seems to decrease visualization of ghrelin-immunoreactive cells, occasional cells containing both ghrelin immunoreactivity and GOAT mRNA are still observed, as evidenced by the representative cell magnified in the inset (B′). Scale bars, 50 μm in A and B, 5 μm in B′.

Expression of GOAT and other MBOATs within a ghrelin cell-enriched pool of gastric mucosal cells.

We prepared highly enriched pools of gastric ghrelin cells by taking advantage of the humanized Renilla reniformis green fluorescent protein (hrGFP) reporter present in a recently characterized line of transgenic mice in which hrGFP is expressed under the control of ghrelin-regulatory elements contained within a bacterial artificial chromosome (20). We previously had shown for this particular ghrelin-hrGFP line (hrGFP10) that 100% of the hrGFP-containing gastric mucosal cells coexpressed ghrelin and that nearly 99% of ghrelin-immunoreactive gastric cells contained hrGFP fluorescence (20). Cells comprising the gastric mucosa of the hrGFP10 line were first enzymatically and mechanically dispersed and subsequently submitted for FACS analysis to generate an enriched sample of hrGFP-positive cells. For each of four independent preparations, those cells with the most intense fluorescence were collected as the hrGFP-positive pool, whereas those with the least intense fluorescence were collected as the hrGFP-negative pool (Fig. 3, top).

Fig. 3.

Relative expression of various mRNAs within 4 different preparations of FACS-separated hrGFP-positive vs. hrGFP-negative gastric mucosal cells. Top: graphical representation of the FACS of gastric mucosal cells from 1 of the 4 different preparations, indicating those cells collected as part of the hrGFP-enriched pools and those collected as part of the hrGFP-negative pools. Bottom: quantitative RT-PCR data demonstrating relative levels of expression of mRNAs encoding ghrelin, 2 markers of gastric mucosal cell types, GOAT, and other MBOAT (membrane-bound O-acyltransferases) family members (vs. mRNA levels of the housekeeping gene cyclophilin).

Next, quantitative RT-PCR was performed on mRNAs isolated from both the hrGFP-positive and the hrGFP-negative pools, for each of the four independent preparations, and the average levels of various mRNA species relative to that of the housekeeping gene cyclophilin were determined (Fig. 3, bottom). Although ghrelin could be detected in small amounts within the hrGFP-negative pools of gastric mucosal cells (most likely as the result of incomplete expression of hrGFP within all ghrelin cells and slight incompleteness in the sorting), ghrelin mRNA expression was found to be significantly higher within the hrGFP-positive pools, thus confirming the success of the fluorescence-activated cell sorting. The mostly endocrine nature of these hrGFP-positive pools was evidenced by the relatively high levels of chromogranin A compared with the hrGFP-negative pools. Pepsinogen F, a known marker for the much more common parietal and zymogenic gastric mucosal cells, was also detected at very low levels within the hrGFP-enriched pools, although its levels were statistically higher in the hrGFP-negative pools. Importantly, although present at much lower levels than ghrelin mRNA, GOAT mRNA was also highly enriched for within the hrGFP-positive pool, thus reaffirming the earlier dual-label histochemistry results. We also found evidence for all the other members of the MBOAT family within the hrGFP-positive pools, albeit at levels far below those of GOAT. Of the MBOAT family members, only HHAT and ACAT2 were found to meet criteria for being present at statistically significantly higher levels within the hrGFP-positive pools, although the functional significance of these increases is unclear given the very low levels.

Tissue distribution of MBOAT family members.

RT-PCR analysis of cDNAs prepared from fifteen different murine tissues was performed to assess the distribution of the mRNAs encoding all the MBOAT family members; for comparative reasons, similar analyses were performed for preproghrelin (Fig. 4). As shown previously, preproghrelin mRNA was expressed most highly in the stomach and along the remainder of the gastrointestinal tract (29). Also as previously shown, low levels of preproghrelin mRNA were present in total brain, hypothalamus and testis (29). Notably, the expression pattern of GOAT mRNA matched that of preproghrelin nearly perfectly - including similarities in the tissues with expression (with the exception of kidney) and the relative expression levels within those tissues. None of the other MBOAT family members demonstrated expression patterns matching that of ghrelin, and in fact, most were found in all of the tissues examined.

Fig. 4.

Tissue distribution of preproghrelin, GOAT, and other MBOAT family members. Relative levels, within a series of 15 difference mouse tissues, of mRNAs encoding preproghrelin and the 16 members of the MBOAT family. The acyl acceptor substrates for each MBOAT member are listed within the parentheses.


In this paper, we describe for the first time a high degree of colocalization of GOAT and ghrelin within the same gastric mucosal cells. This was demonstrated using multiple approaches, including both dual-label histochemical analysis of mouse stomach sections and quantitative RT-PCR on mRNAs from a highly enriched pool of mouse gastric ghrelin cells. Coexpression of ghrelin and GOAT was also observed within duodenal mucosal cells. As ghrelin is the only known acyl substrate for GOAT, this high degree of colocalization was not entirely unexpected. However, in the original studies of GOAT, it remained an open question whether GOAT was functioning inside ghrelin cells or whether GOAT residing in adjacent non-ghrelin cells might “trans”-modify desacyl-ghrelin secreted by ghrelin cells. Our current findings support the notion that ghrelin octanoylation happens during its processing procedure, within gastric ghrelin cells, and should help focus further investigations into the regulation of the posttranslational processing of ghrelin.

We also observed smaller numbers of gastric ghrelin cells that did not seem to coexpress GOAT. Within the duodenum, the percentage of these presumed ghrelin-immunoreactive cells lacking GOAT was even higher. The most straightforward explanation for these findings is that there exists a subset of gastric ghrelin cells that do not produce octanoylated ghrelin and that the majority of duodenal ghrelin cells do not produce octanoylated ghrelin. The differences observed between the stomach and the duodenum potentially could be related to heterogeneity among ghrelin cells within those two organs. For instance, within the stomach, ghrelin cells are thought to be only of the closed type, which do not have direct access to stomach luminal contents (19). However, the duodenal ghrelin cell population includes both the closed type and the open type, which do interact directly with luminal contents (19). Thus, although speculative, it could be that only one of those types of ghrelin cells contains GOAT (for instance, the closed type) and that fewer of the duodenal ghrelin cells contain GOAT because the duodenum contains a lower percentage of closed-type ghrelin cells. The implication of ghrelin cells lacking GOAT is unclear at this time, although there have been several reports that suggest a role for des-acyl ghrelin in certain physiological processes, such as regulation of blood pressure and cell proliferation (4, 5, 25).

Within the stomach, we also observed a small percentage of presumed GOAT mRNA-containing gastric cells that did not contain ghrelin immunoreactivity. This suggests that there may be other substrates besides ghrelin for GOAT. The existence of alternative substrates for GOAT could potentially limit the use of any targeted therapy against GOAT as a means of neutralizing acylated ghrelin, although such alternative substrates have not been identified to date.

Importantly, at least some, if not many, of the presumed GOAT-only gastric mucosal cells (lacking ghrelin immunoreactivity) and the presumed ghrelin-only (lacking GOAT mRNA) gastric mucosal cells might be the result of incomplete labeling. Such is considered a technical limitation when dual-label histochemistry is performed, especially for mRNA species that have very low levels of expression, such as GOAT. Not only do the low levels of GOAT mRNAs (as confirmed by quantitative PCR) pose problems, but it also is often the case that combining methods to histologically detect two different targets on the same tissue section can reduce the efficiency of detection of one or both of the targets. As such, the observed number of dual-labeled cells may reflect a much larger, actual number of dual-labeled cells that lie beyond the limit of detection especially when very stringent parameters for defining dual labeling are used.

Given the high coexpression of GOAT and ghrelin within the stomach, it is likely that, within the other tissues in which we observed GOAT mRNA by RT-PCR, such as the duodenum, jejunum, ileum, colon, and testis, GOAT also occurs within the ghrelin cell populations of those tissues. We confirmed this for duodenum by use of histochemistry, although we would also need to perform dual-label histrochemistry to definitively confirm GOAT/ghrelin coexpression within individual cells comprising those other tissues. Our current description of both GOAT and ghrelin mRNAs within the brain is interesting, since previous descriptions of ghrelin expression within the brain have not been consistent. For instance, several studies report ghrelin mRNA in the brain following RT-PCR analyses (26). However, ghrelin immunoreactivity has not been consistently localized to a single brain region (2, 11, 15, 18) and alternatively has been shown to occur in the hypothalami of ghrelin-deficient mice, thereby suggesting some nonspecificity for at least some of the available anti-ghrelin antisera (28). In our own initial characterization of two different ghrelin-hrGFP-transgenic lines, we did not detect hrGFP within the brain (20). Within human tissues, brain was not one of the tissues in which ghrelin mRNA or GOAT mRNA was detected (6). However, the current finding of GOAT mRNA within total brain and hypothalamic extracts, in addition to ghrelin mRNA, does give credence to the assertion of CNS ghrelin expression, at least in certain species. The physiological significance of central ghrelin remains to be determined.

The high concordance of not only the sites of coexpression but also the relative levels of coexpression of ghrelin and GOAT mRNAs also suggests a common transcriptional regulation of the two genes. Future analysis of any conserved DNA sequence(s) within the promoter regions of ghrelin and GOAT will be an interesting option to uncover those transcription factors controlling ghrelin and GOAT expression in specific cell types. Also, such an analysis may further shed insight on the differentiation of ghrelin-producing neuroendocrine cells during mammalian development.

Our current study also demonstrates the distribution of all of the other MBOAT mRNAs besides GOAT in a variety of mouse tissues. Similar to GOAT, both HHAT and PORC(a/b/c/d) acylate proteins [Sonic Hedgehog and Wnt, respectively (1, 27)]. A large body of evidence suggests that Hedgehog and Wnt not only are important morphogens functioning in animal development to guide cell differentiation but also play critical roles in maintenance of adult stem cells to control the balance of cellular death and proliferation or regeneration after wounds in adult tissues (8, 10). Our finding of universal expression of HHAT and PORC mRNAs within all the tested mouse tissues seems consistent with these roles. We also observed ubiquitous expression of ACAT1, whereas ACAT2 was localized only to the gastrointestinal tract and the liver, which agrees with previous findings (14). Although mRNAs encoding HHAT and ACAT2 were both found to be statistically enriched within the FACS-separated gastric mucosal ghrelin (hrGFP-positive) cells compared with the hrGFP-negative gastric mucosal cells, the functional and physiological significance of this finding remains uncertain, especially as it relates to ghrelin, since neither is known to posttranslationally modify ghrelin (29) and since their levels are so low, especially compared with GOAT.

Recently, the substrates for the other previously considered orphan MBOAT family members have been characterized, including GUP1, LRC4, MBOAT1, MBOAT2, and MBOAT5 (7, 9, 13, 22). Intriguingly, those MBOAT members all function as glycerophospholipid-remodeling enzymes or lysophospholipid acyltransferases. In our expression survey, LRC4 and MBOAT5 are detectable in most of the mouse tissues, with only minor fluctuations in expression levels, whereas GUP1, MBOAT1, and MBOAT2 show much more individualized expression patterns. The absence of lysophospholipid acyltransferases in some tissues, for instance, little or no GUP1 mRNA in gastrointestinal tract, may indicate that cells in those tissues possess a unique composition or metabolize only certain phospholipids to accommodate their physiological functions.

In summary, we have shown a high degree of localization of GOAT mRNA specifically to ghrelin cells within the stomach as well as a high concordance of GOAT and ghrelin expression throughout several different tissues. GOAT's other family members, which previously were shown to lack the ability to acylate ghrelin, have tissue distribution patterns that do not emulate those of ghrelin or GOAT. Our data support a role for the gastric ghrelin-producing cell as the key site for ghrelin's important posttranslational acylation with octanoate by GOAT.


This work was supported by funding from the National Institutes of Health (1F32-DK-64564-01, K08-DK-068069-01A2, R01-DK-71320, HL-20948, 1RL1-DK-081185-01, 1PL1-DK-081182-02 and DK-56116), a grant from the Perot Family Foundation, and a UTSW Medical Center Disease-Oriented Clinical Scholars Award.


We acknowledge the support, guidance, and advice of Drs. Michael S. Brown and Joseph L. Goldstein. We also thank Dr. Guanghua Xiao from University of Texas Southwestern (UTSW) Medical Center's Department of Clinical Sciences, Division of Biostatistics, for advice on the statistical analyses.


  • * I. Sakata and J. Yang contributed equally as first authors.


View Abstract