Biosynthesis of Phytosphingosine by the Rat*

The co-migration on gas-liquid chromatography of the trimethylsilyl ethers of the free and borohydride (d) obtainment same fragmentation

The criteria for identification were: (a) co-migration on gas-liquid chromatography of the trimethylsilyl ethers of the free and N-acetylated base with authentic standard, (b) co-migration on thin layer chromatography of the dinitrophenyl derivative with authentic standard, (c) labeled pentadecanol detection by gas-liquid radiochromatography after subjecting intact aphingolipids or a mixture of free long chain bases to periodate oxidation followed by sodium borohydride reduction, and (d) obtainment of the same fragmentation pattern as with authentic standard when the trimethylsilyl ethers of the N-acetylated base were subjected to gas-liquid chromatography-mass spectrometry. Smaller amounts of labeled C,,-phytosphingosine were also detected in eeramide-and sphingomyelin-rich fractions of intestine and in liver and kidney. Experiments with germfree rats strongly suggest that the intestinal flora is not the (sole) site for the biosynthesis of C,,-phytosphingosine.
Phytosphingosines are long chain bases that differ from sphingosines in that they have an additional hydroxyl at C-4 and have no double bond between carbon 4 and carbon 5. They were first isolated by Carter and collaborators in plants (0, hence their name "phyte"-sphingosines.
The configuration of the amino and hydroxyl groups was determined by Carter and Hendrickson (2) to be n-ribo-1,3,4-trihydroxy-2-aminooctadecane. These materials and their unsaturated and branched analogs have since been reported to occur in yeasts (3), protozoa (4,5). sea urchin (6), fish (7), amoeba (8), and mammals where they have been detected as part of sphingolipids in human (91, bovine (10, ll), rat (11,12) and equine (13) kidney, rat intestine (14, 151, bovine milk (161, porcine erythrocytes (17), and in one case of adenocarcinoma (18 We reasoned that if labeled C,,-phytosphingosine occurred in such fraction, then according to the scheme of Fig. 1, one should be able to isolate and identify labeled pentadecanol as a reaction product upon periodate oxidation followed by sodium borohydride reduction. Fig. 2 shows a gas-liquid radiochromatogram of such a sample obtained 24 h atier the injection of labeled C,,-dihydrosphingosine. The predominant mass peak and most of the radioactivity had the same mobility as pentadecanol. A similar mass profile was obtained from a sample isolated 4 h aRer injection of labeled C,,-dihydrosphingosine (Fig. 3) 3. Gas-liquid radiochromatographic analysis of long chain alcohols obtained after periodate oxidation followed by sodium borohydride reduction of a sphingolipid-rich fraction of intestine from rats 4 h after injection of D-e@hrO[4,5-SHIC,,-dihydrosphingosine.
A, pentadecanol; B, hexadecanol. and eluting the column with chloroform, chloroform:methanol (97:3) and chloroform:methanol (60:40). Almost half of the radioactivity eluted in this latter fraction from a sample obtained 24 h after injection of labeled C,,-dihydrosphingosine co-migrated on thin layer chromatography with bovine brain cerebrosides. The remainder of the radioactivity had a mobility consistent with ceramide polyhexosides.
No labeled free long chain bases could be detected. Periodate oxidation of this fraction followed by sodium borohydride reduction yielded pentadecanol as the only long chain alcohol detectable by gasliquid chromatography and virtually all the radioactivity comigrated with this peak (data not shown).
When this glycolipid-rich fraction was subjected to acid hydrolysis and the resulting free long chain bases were separated by gas-liquid chromatography (as the Me,Si ethers), the profile shown in Fig. 4 was obtained. Two major mass peaks can be seen. The first to elute (33% of the total peak areas) had the same mobility as the Me,Si ethers of Clx-sphingosine.
The second peak (67% of the total peak areas) co-migrated with the Me,Si ethers of C,,-phytosphingosine. This same figure shows that a substantial portion of the radioactivity (-33%) comigrated with the Me,Si ethers of C,,-phytosphingosine while the remainder co-migrated with the Me,Si ethers of CIH-sphingosine and C,,-dihydrosphingosine.
A similar analysis of the sample obtained 4 h affer injection of radiolabeled C,,-dihydrosphingosine showed that only 6% of the radioactivity co-migrated with the Me,Si ethers of C,"phytosphingosine.
The remainder co-migrated with the Me,Si ethers of C,,-dihydrosphingosine.
Because of the relative small amount of labeled C,,-phytosphingosine in this sample, all further characterizations of the labeled C,"-phytosphingosine were done with the sample obtained 24 h after injection of labeled Cl"-dihydrosphingosine.
To determine whether the mass peak which co-migrated with the Me,Si ethers of C,,-phytosphingosine on Fig. 4 was homogeneous, the mixture of free long chain bases was converted to the N-acetyl derivative and subsequently subjected (as the Me,Si ethers) to gas chromatography-mass spectrometry analyses. The mass fragmentation patterns on both sides of the gas chromatograph peak that had been tentatively identified as the Me,Si ethers of N-acetyl-C,,-phytosphingosine were determined and found to be very similar, arguing against a mixture ofiV-acetyl-C,,-phytosphingosine with some other unknown compound. The following fragment ions that have been reported to be characteristic for the Me,Si ether derivative of N-acetyl-C,,-phytosphingosine were observed: m/e 299 (20% of base peak); m/e 560 (M-15), and m/e 401. The base peak was mle 332 or mle 73, in agreement with the fragmentation patterns previously described (18,23,25). The free long chain bases were also converted to the dinitrophenyl derivatives.
Upon separation by thin layer chromatography, the radioactivity profile shown in Fig. 5 was obtained. Approximately 30% of the radioactivity that co-migrated with the DNP derivatives of long chain bases had the same mobility as DNP C,,-phytosphingosine.
The remainder co-chromatographed with DNP derivatives of C,,-sphingosine and C,,dihydrosphingosine.
The radioactivity at the origin was not characterized.
The mixture of free long chain bases was also oxidized with periodate and the resulting long chain aldehydes were reduced with sodium borohydride and the alcohols analyzed by gasliquid chromatography (Fig. 6). The principal mass peaks comigrated with pentadecanol and AZ-hexadecenol. Thirty five of Phytosphingosine by the Rat per cent of the radioactivity co-migrated with pentadecanol, as to be expected if it was derived from C,,-phytosphingosine ( Figs. 1 and 4). The remainder of the radioactivity co-migrated with hexadecanol and A*-hexadecenol, in the percentages to be expected if it was derived from C,,-dihydrosphingosine and C,,-sphingosine ( Figs. 1 and 4).
Occurrence of Labeled C,!,-Phytosphingosine in Other Tissues and Lipid Fractions -Labeled C,,-phytosphingosine was also detected in ceramide-rich and sphingomyelin-rich fractions of intestine and in the corresponding ones of kidney and liver. However, in all these samples, the radioactivity was less than 8% of that in other long chain bases (C,*-sphingosine + C,,-dihydrosphingosine).

Experiments
with Germ-free Rats-We had previously postulated (11) that the intestinal flora did not play a significant role in the biosynthesis of phytosphingosine.
This hypothesis was tested at the University of Notre Dame by using germ-free rats from their colony.
Male rats (1 year old, 250 g average weight) were injected with sterile 14,5-SH]C,,-dihydrosphingosine (0.55 mCi in 5% propylene glycol in saline). Animals were killed 24 h after injection. Upon periodate oxidation (followed by sodium borohydride reduction) of a crude sphingolipid fraction of rat intestine, 60% of the radioactivity in long chain alcohols was in pentadecanol, strongly suggesting that C,,-phytosphingosine was not (solely) derived from the intestinal flora. DISCUSSION Assmann and Stoffel (26) observed in a previous study that radiolabeled C,,-phytosphingosine was incorporated into sphingolipids by the rat when injected intravenously or administered orally. These experiments led them to conclude that phytosphingosines in rats were of dietary origin. Our results with normal and germ-free animals clearly show that the rat can synthesize phytosphingosines and strongly suggest that the diet or the intestinal flora (or both) are not the sole source of phytosphingosines in mammals. Several studies have suggested that C,,-phytosphingosine originates by hydroxylation of C,,-dihydrosphingosine (23,24,27). Although, in the present study, labeled C,,-phytosphingosine may be directly derived from C18-dihydrosphingosine, it is also possible that labeled C,,-dihydrosphingosine was first broken down to [2,3-3H]palmitie acid and that subsequently this material was incorporated into phytosphingosine by some other mechanism. Furthermore, since the amount of label that was actually isolated and characterized as C&phytosphingosine was less than 0.1% of the radioactivity injected, we cannot rule out the possibility that the precursor of C,,-phytosphingosine was a small quantity of an unknown contaminant.
The possibility that a small amount of labeled C,,-phytosphingosine (as a contaminant of the C,,-dihydrosphingosine) was actually injected into rats and subsequently concentrated by different tissues is highly unlikely, as preparative thin layer chromatography (used in the purification) clearly separates C,,,-dihydrosphingosine from C,,-phytosphingosine; in addition, this latter base cannot be purified by this procedure because it breaks down during elution from the silica gel.
We have found that periodate oxidation of intact sphingolipids followed by sodium borohydride reduction is a rather sensitive method to detect C,,-phytosphingosine (in the form of pentadecanol).
This procedure has enabled us to detect a mass peak for pentadecanol in only those fractions where subsequently C,,-phytosphingosine was detected as the Me,Si ethers of the free base. The possibility that the labeled penta-decanol was not solely derived from C,,-phytosphingosine but also from cY-hydroxypalmitate of high specific activity (bound to the amino group of the long chain bases) was also investigated. Essentially no radioactivity was detected in this fatty acid.
C,,-phytosphingosine (in small amounts) has been found to be rather unstable to chemical treatments in which other long chain bases such as C18-sphingosine and C,,-dihydrosphingosine are relatively stable. For example, as previously mentioned, we have been unable to purify C,,-phytosphingosine by preparative thin layer chromatography.
We have also detected breakdown of C,,-phytosphingosine during acid hydrolysis of small samples of sphingolipids under conditions where CIHsphingosine and C,8-dihydrosphingosine are rather stable. Variations of the percentage of radiolabeled C,,-phytosphingosine (as compared to other long chain bases) in a given tissue of rats injected with the same amounts of radiolabeled CIHdihydrosphingosine have also been observed. Some of this variability can be attributed to individual differences in metabolism; however, this problem is further complicated by the previously mentioned instability of small amounts of C,,-phytosphingosine isolated during such studies. One therefore should keep in mind that the 33% of radioactivity in the long chain bases that was characterized as C,,-phytosphingosine of a glycolipid-rich fraction of intestine from animals 24 h after injection of labeled C,,-dihydrosphingosine represents an average from four rats.
The occurrence of phytosphingosine mass in some sphingolipids but not others, even within the same tissue, raises some intriguing questions as to how this base is incorporated into sphingolipids and its relation to other long chain bases. We are currently attempting to gain further insights into some of these questions.