Enzymes involved in L-carnitine biosynthesis are expressed by small intestinal enterocytes in mice: Implications for gut health

Abstract

Background

Carnitine is essential for mitochondrial β-oxidation of long-chain fatty acids. Deficiency of carnitine leads to severe gut atrophy, ulceration and inflammation in animal models of carnitine deficiency. Genetic studies in large populations have linked mutations in the carnitine transporters OCTN1 and OCTN2 with Crohn’s disease (CD), while other studies at the same time have failed to show a similar association and report normal serum carnitine levels in CD patients.

Methods

In this report, we have studied the expression of carnitine-synthesizing enzymes in intestinal epithelial cells to determine the capability of these cells to synthesize carnitine de novo. We studied expression of five enzymes involved in carnitine biosynthesis, namely 6-N-trimethyllysine dioxygenase (TMLD), 4-trimethylaminobutyraldehyde dehydrogenase (TMABADH), serine hydroxymethyltransferase 1 & 2 (SHMT1 & 2) and γ-butyrobetaine hydroxylase (BBH) by real-time PCR in mice (C3H strain). We also measured activity of γ-BBH in the intestine using an ex vivo assay and localized its expression by in situ hybridization.

Results

Our investigations show that mouse intestinal epithelium expresses all five enzymes required for de novo carnitine biosynthesis; the expression is localized mainly in villous surface epithelial cells throughout the intestine. The final rate-limiting enzyme γ-BBH is highly active in the small intestine; its activity was 9.7 ± 3.5 pmol/mg/min, compared to 22.7 ± 7.3 pmol/mg/min in the liver.

Conclusions

We conclude that mouse gut epithelium is able to synthesize carnitine de novo. This capacity to synthesize carnitine in the intestine may play an important role in gut health and can help explain lack of clinical carnitine deficiency signs in subjects with mutations with OCTN transporters.

Introduction

Carnitine (β-hydroxyl-γ-trimethylaminobutyric acid) acts as an obligatory cofactor for β-oxidation of fatty acids by facilitating transport of long-chain fatty acids across the inner mitochondrial membrane as acylcarnitine esters. Carnitine plays a major role in tissues (e.g., heart, skeletal muscle, liver, placenta, and the intestinal tract), which depend on β-oxidation of fatty acids for energy production. Over >99% of carnitine in the body is intracellular, and the organic cation transporter OCTN2 (SLC22A5) is primarily responsible for maintaining this tissue gradient ¹. Two distinct types of carnitine deficiencies have been identified. Primary carnitine deficiency arises from defects in the plasma membrane carnitine transporter OCTN2. Patients with this disorder excrete carnitine in urine due to defective reabsorption and plasma and tissue levels of carnitine may drop below 10% of normal values ^2–6. Secondary carnitine deficiency arises from defects in any of the enzymes involved in mitochondrial fatty acid oxidation where organic acids accumulate due to block in this metabolic pathway; these organic acids then bind to carnitine and are excreted in urine in the form of acylcarnitines ^{7, 8}. It has been estimated that roughly 75% of total carnitine in the body is obtained from dietary sources and about 25% comes from endogenous synthesis ⁹.

Carnitine is biosynthesized from essential amino acids lysine and methionine (Figure 1). ε-N-trimethyllysine, which is formed by methylation of protein-bound lysine and is released in free form after protein degradation, is the precursor for carnitine biosynthesis. This lysine derivative is hydoxylated by ε-N-trimethyllysine dioxygenase (TMLD). Subsequently, the resulting β hydroxyl-ε-N-trimethyllysine is cleaved into γ-trimethylaminobutyraldehyde and glycine by β-hydroxy-ε-N-trimethyllysine aldolase (HTMLA), after which the aldehyde is oxidized by γ-trimethylaminobutyraldehyde dehydrogenase (TMABADH) to yield γ-butyrobetaine. The final step in this pathway involves the hydroxylation of γ-butyrobetaine by γ-butyobetaine hydroxylase (γ-BBH) to yield L-carnitine ^{10, 11}.

Figure 1. Pathway of de novo L-carnitine biosynthesis.

Carnitine is biosynthesized from TML following its release from proteins by degradation. TML is hydroxylated by TMLD resulting in production of HTML, which is further cleaved to TMABA by an aldolase in the presence of pyridoxal 5′-phosphate. Subsequently, TMABA is oxidized to γ-BBH by TMABA-DH and in the final step γ-BBH is hydroxylated to carnitine by γ-BBH.

While clinical consequences of carnitine deficiency (primary or secondary) in the myocardium and skeletal muscle has received much attention, the relevance of carnitine deficiency to intestinal tract, particularly during inflammation has recently been recognized and it may thus have a clinical role in gut health ^12–16. The role of carnitine in the intestinal tract has been highlighted by several publications linking mutations in genes encoding carnitine transporters OCTN1 (SLC22A4) and OCTN2 (SLC22A5) with Crohn’s disease (CD) ^17–23. Patients with CD have been shown to have a missense substitution 1762C→T in OCTN1 which causes amino acid substitution L503F and a G→C transversion in the promoter region of OCTN2 (-207G→C), which disrupts a heat shock binding element (HSE) in the promoter region of OCTN2 gene. These mutations have been shown to decrease plasma membrane transport of carnitine in transiently transfected cell line studies and may thus reduce tissue content of carnitine¹⁷. Over the past several years, these human genetic studies have been replicated in large populations with CD, suggesting a cause-and-effect relationship between these mutations and inflammatory bowel disease. A meta-analysis of 12 of these studies with >3000 subjects in each arm showed a robust association between these mutations and CD ²⁴. Therefore, carnitine supplementation could potentially have a therapeutic benefit in patients with inflammatory bowel disease (IBD) in general and CD in particular. However, these studies are in sharp contrast with few other reports, which have failed to show a clear link among plasma levels of carnitine, CD and OCTN1 (1762C→T) and OCTN2 (-207G→C) mutations ^{25, 26}. This discrepancy may be explained by the diffrences between plasma and tissue carnitine levels since tissue levels were not measured in human subjects with IBD. At the same time a few recent reports in animals ^{27, 28} and humans ²⁹ have demonstrated beneficial effects of carnitine supplementation on gut inflmmatory markers and patient symptoms.

In view of this controversy regarding any potential releationship between defective carnitine transporters and IBD, we asked whether the intestinal tract is capable of synthesizing carnitine de novo. If this tissue does indeed possess the enzymatic machinery necessary to synthesize carnitine, it might explain the lack of gut pathology in patients with inactivating mutations in the carnitine transporters. We addressed this question by examining the expression of enzymes involved in de novo biosynthesis of carnitine in mouse intestinal tract.

Methods

Animals and sample preparation

Six normal mice (C3H strain) were used in these studies. Tissues from various parts of intestinal tract, and other viscera were thoroughly washed and freed from blood, were placed in chilled phosphate-buffered saline (PBS) and snap-frozen at −80°C for further analysis. All experimental procedures were performed with approval from the Institutional Animal Care and Use Committee at the Georgia Health Sciences University.

Quantitative Real-time -PCR for enzymes involved in carnitine biosynthesis

Quantitative PCR was performed to study expression of five enzymes involved in carnitine biosynthesis, these studies produced amplicons of 168 to 187 bp size. Primers were designed using IDT primer quest software available online (Integrated DNA technologies, Skokie, IL, USA). The genes encoding TMLD, TMABADH and γ-BBH have been cloned and characterized in mouse, but the second enzyme of this pathway HTMLA, a pyridoxal 5′-phosphate-dependent aldolase, has not been identified. It has been speculated that HTMLA is identical to two other enzymes, a serine hydroxymethyltransferase 1 & 2 (SHMT1 & 2) based on the observations that purified SHMTs are able to convert HTML to TMABA ^{30, 31}. Therefore, we designed primers for SHMT-1 (accession No. NM_009171) and SHMT-2 (accession No. BC_051396), both of which have been shown to carry similar function. The nucleotide sequences of the primers used for realtime-PCR studies and sizes of the PCR products are as shown in Table 1. One μg of the total RNA from each tissue sample was reverse transcribed with random hexamers and reagents from RNA PCR kit (Perkin Elmer, Norwalk, CT, USA). Real-time PCR was performed using a BioRad iQ iCycler Detection System (BioRad Laboratories, Hercules, CA, USA) with SYBR green fluorophore. Reactions were performed in a total volume of 25 μL, which included 12.5 μL 2x SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA, USA), 1 μL of each primer at 5μM concentration, and 1 μL of the previously reverse-transcribed cDNA template and 9.5 μL of dH₂O. Thermal cycling conditions were as follows: an initial incubation at 95 °C for 10 min was followed by 50 cycles of 95 °C for 30 s, 60 °C for 45 s and 72 °C for 45 s, and a final incubation at 72 °C for 5 min. The concentration of gene-specific mRNA in tissue samples was calculated as described earlier³². Each reaction was carried out in duplicate from three animals for each type of tissue sample (n=6 in each group), data analysis was carried out by proprietary software supplied by the vendor.

Table 1.

Primers used for Real-time PCR to studies

Accession Number	Enzyme	Amplicon (bp)	Sense primer	Antisense primer
NM_138758	TMLD	177	TGG CAC TGT GAT GCG CTT TGA T	TGA CCA TCT GGC CAA GTG AA
NM_019993	TMABADH	175	TGA TTG CCA CTT TCG CGT GT	ATG GTC TCC ACG GTG GCA ATT T
NM_009171	SHMT-1	185	ACA CTG CAG ATT CAG AGC CAC A	TTG GCA AAC ACA GGC TGT TCC T
BC_051396	SHMT-2	168	ACG CGT GTT GGA ACT TGT CT	TCC AAG CCA ATG TTG ACT CCC T
NM_130452	γ-BBH	187	AGT GGT TCG CGT CAA CTT CA	ATG AAG CAA GCG CCA GTT GT
NM_013556	HPRT	176	GCG TCG TGA TTA GCG ATG ATG AAC	CCT CCC ATC TCC TTC ATG ACA TCT

γ-Butyrobetaine hydroxylase activity assay

γ-Butyrobetaine hydroxylase is the key terminal rate-limiting enzyme in the pathway for biosynthesis of carnitine. We measured its activity in various mouse tissues in comparison to the epithelial scrapings from small intestine. γ-Butyrobetaine is hydroxylated by this enzyme to generate carnitine, which was then measured as described earlier ³³. The assay reaction was carried out in 1.5 ml Eppendorf centrifuge tubes in a volume of 0.2 ml containing 10 mM TES [N-{(trishydroxymethyl)methyl}-2-aminoethane-sulfonic acid] buffer (pH 7.2), 200 μM ferrous sulfate, 2 mM ascorbate, 4 mM α-ketoglutarate, catalase, and 10 μg protein from various tissues. The reaction was initiated by addition of γ-butyrobetaine (10 mM), and the mixture was incubated in a water bath at 37°C for 30 min. Reaction was terminated by boiling the reaction mix at 100°C for 5 min.

Tissue carnitine measurement

Fifty to 100 mg of epithelial scrapings from jejunum, ileum and colon, along with liver, heart and skeletal muscle from all six animals was placed in PBS and homogenized using a Polytron homogenizer. The protein concentration of the homogenate was measured by the Lowry method. Total, free and acyl carnitine fractions in various tissue samples was determined by tandem mass spectrometry as described ^{34, 35}.

In situ hybridization for γ-butyrobetaine hydroxylase gene expression

Fresh mouse intestinal tissues were fixed in OCT media (Tissue Tek #4583) and snap-frozen at −80°C. Tissue sections were prepared by the Histology Core facility. The expression of BBH gene (accession number NM_130452) in these tissue sections was studied by in situ hybridization. For the preparation of antisense and sense (negative control) riboprobes, a 585 bp EcoRI fragment, which corresponds to nt 354-939 of the γ-butyrobetaine hydroxylase cDNA, was cloned into pGEM-T vector. The probes were prepared by in vitro transcription using the DIG nucleic acid detection kit with appropriate RNA polymerases (T7 RNA polymerase for the sense probe and SP6 RNA polymerase for the antisense probe) after linearizing the plasmid with SalI for the sense probe and Apa1 for the antisense probe. In situ hybridization was performed as described in the kit (Roche Diagnostics, Indianapolis, IN).

Statistical Analyses

γ-BBH enzyme activity was measured in six different samples for each tissue. Data are presented as means ± SD and comparisons between paired samples were made by Student’s ‘t’ test with Bonferroni’s correction where applicable and statistical significance was set at p ≤ 0.05. We used statistical software SPSS for PC version 11.01 for data analysis.

Results

Mouse intestinal tract expresses all five enzymes required for de novo carnitine biosynthesis

Real-time PCR studies (Figure 2) showed that the expression of genes of two enzymes, TMLD and TMABADH, in the small intestine was comparable or even higher to that in the liver, the organ recognized as the primary site of endogenous carnitine biosynthesis. The other three enzymes, SHMT1, SHMT2, and γ-BBH, were also expressed in the small intestine, though less than in the liver and kidney. Kidney expressed genes of four of the five enzymes in robust amounts, indicating that this organ also possesses the capacity for carnitine biosynthesis. In contrast, skeletal muscle did not express genes of four enzymes (SHMT1, SHMT2, TMABADH and γ-BBH) in any appreciable amount. Thus this tissue is not likely to generate carnitine endogenously and hence has to depend on carnitine in circulation for its fatty acid oxidation needs.

Figure 2. Expression of enzymes involved in carnitine biosynthesis in mice.

A composite image of real-time-PCR experiments showing relative mRNA expression levels specific for TMLD, TMABADH, SHMT1, SHMT2 and γ-BBH in small intestine, liver, kidney, and skeletal muscle in relation to expression of HPRT.

Carnitine content in intestinal and colonic epithelium is comparable to that in other tissues

Total carnitine (n=6) content in epithelial scrapings of the small intestine (jejunum and ileum) and colon was comparable to that in liver, heart, and skeletal muscle (Table 2), the organs which are known for robust fatty acid oxidation as an important source of metabolic energy.

TABLE 2.

Total tissue carnitine content (nmol/mg) of gut scrapings in relation to other organs (n=6)

Tissue	Jejunum	Ileum	Colon	Liver	Heart	Skeletal muscle
Total carnitine (nmol/mg)	7.0 ± 1.2	8.8 ± 3.0	7.1 ± 0.3	6.9 ± 2.8	8.2 ± 0.9	5.4 ± 1.9

Mouse small intestinal epithelium possesses robust γ-BBH activity

We measured the functional activity of γ-BBH, the rate-limiting terminal enzyme of the carnitine biosynthetic pathway, in mucosal scrapings of mouse small intestine and compared the activity with that found in other tissues. Liver had the highest activity, followed by placenta. Small intestine had appreciable activity, amounting to 43% of that in liver (Figure 3). Both Liver and placenta have been shown to synthesize carnitine ^{10, 36}. The activity of this enzyme was much less in heart and skeletal muscle.

Figure 3. γ-butyrobetaine hydroxylase (γ-BBH) activity in mouse small intestine.

Gut scrapings from mouse small intestine had γ-BBH activity of 9.7 ± 3.5 pmol/mg/min, in comparison to that of liver 22.7 ± 7.6 pmol/mg/min.

γ-BBH expression is localized mainly to the luminal surface epithelial cells throughout the intestinal tract

We examined the location of γ-BBH expression in the intestinal tract by detecting its mRNA in tissue sections by in situ hybridization (Figure 4, 5 & 6). These studies showed intense staining for γ-BBH mRNA in surface epithelial cells in the small intestine (jejunum and ileum) as well as in the colon. There was minimal staining in other cell types of the intestinal tract. The observed hybridization signals were specific for γ-BBH mRNA; the sense probe did not yield any signal in the tissue sections examined.

Figure 4. High-resolution image (×20) of in situ hybridization for γ-BBH expression in mouse jejunum.

Composite image of γ-BBH expression in jejunum; panel A is hybridized with an antisense probe and panel B with a sense probe. Panel A show dark staining localized mainly to surface epithelial cells (Bar represents 50 μm).

Figure 5. High-resolution image (×20) of in situ hybridization for γ-BBH in mouse ileum.

Composite image to show localization of γ-BBH expression; panel A is hybridized with an antisense probe and panel B with a sense probe. γ-BBH gene expression was seen mainly in the surface epithelial cells of villous structure with faint staining in the crypts and mesenchymal cells of the villi (Bar represents 50 μm).

Figure 6. High-resolution image (×20) of in situ hybridization for γ-BBH in mouse colon.

Composite image to show localization of γ-BBH expression; panel A is hybridized with an antisense probe and panel B with a sense probe. γ-BBH gene expression was seen mainly in the surface epithelial cells of with very faint staining of mesenchymal mucus producing cells (Bar represents 50 μm).

Discussion

Our results demonstrate for the first time that the enzymes involved in the carnitine biosynthetic pathway for de novo production of carnitine are highly expressed throughout the intestinal tract in mouse. Epithelial scrapings from the small intestine are able to convert γ-butyrobetaine to carnitine, showing a robust activity of γ-BBH in comparison to liver. These studies show not only that the intestinal tract expresses the enzymes detectable at the mRNA level but also that the expressed enzymes are functional at the catalytic level. The villous epithelial cell carnitine content is comparable to other organs such as the liver, the organ known to be the primary site of de novo synthesis of carnitine.

These findings have profound implications to gut health because of the obligatory role played by carnitine in gut energy production. Role of short-chain fatty acids in colon health has been recognized for several years now^37–39 but recent animal data reveal that small intestine enterocytes utilizes long-chain fatty acids which require carnitine for their transport into the mitochondria for subsequent β-oxidation. We have earlier shown that carnitine plays a significant role in energy production in the small intestine and that any perturbation of this important metabolic pathway has disastrous consequences on the structure and function of this tissue ²⁴. The villous epithelial cells are metabolically very active and need to generate ATP for fueling multiple functions such as nutrient transport, maintenance of epithelial barrier and immune function. Our current findings underscore the metabolic need of the intestinal tract to maintain a constant availability of carnitine to carry out β-oxidation of long-chain fatty acids for energy production.

Absorption of dietary carnitine via OCTN2 transporter in the intestinal tract has been well characterized ^{40, 41}. This process is modulated by inflammation ⁴² and PPAR-α agonists ^{43, 44}; thus de novo synthesis of carnitine may be critical under the circumstances where the efficiency of the absorptive process is compromised. Surface epithelium of intestinal tract regenerates at a rapid rate ⁴⁵, a biological process with high demand for metabolic energy. Carnitine is critical to maintain the energy status of these cells, and the ability to synthesize carnitine endogenously would ensure adequate amounts of carnitine necessary for optimal function of these cells. Majority of body carnitine (>99%) is intracellular and OCTN2 deficiency, an autosomal recessive disorder, leads to progressive fall in tissue carnitine; patients with defects in OCTN2 present with progressive cardiomyopathy and skeletal weakness and at times with severe liver dysfunction and metabolic decompensation. Carnitine supplementation in these situations raises the tissue carnitine concentrations to only 5–10% of normal, and this seems sufficient for normal function ⁴⁶. Even though decrease-of-function mutations in the OCTN transporters have been demonstrated in a significant number of patients with IBD and CD, many of these patients do not show any evidence of plasma carnitine deficiency ⁴⁷. This may be accounted for by differences between plasma and tissue carnitine levels since human studies have not reported intestinal tissue carnitine levels. Lack of clinical symptoms could also be explained if carnitine is biosynthesized in the human GI tract then our findings in mice may help explain as to why in spite of decreased transport of dietary carnitine in these patients, serum levels are maintained within acceptable normal range.

Abbreviations

TMLD

6-N-trimethyllysine dioxygenase

TMABA

4-trimethylaminobutyraldehyde

HTMLA

β-hydroxy-ε-N-trimethyllysine aldolase

SHMT1 & 2

serine hydroxymethyltransferase 1 & 2

TMABA-DH

TMABA dehydrogenase

γ-BBH

γ-butyrobetaine hydroxylase

OCTN

organic cation transporter novel

HPRT

hypoxanthine phosphoribosyl transferase

CD

Crohn’s disease

IBD

inflammatory bowel disease