Simple and sensitive multi-sugar-probe gut permeability test by high-performance liquid chromatography with fluorescence labelling.

Enteral intake of a mixture of inert, non-metabolic monosaccharide and disaccharide probes, followed by measurement of their urinary probe ratio, is a well known method to investigate gut permeability. However, most applications lack sensitivity, thus a large amount of especially the disaccharide lactulose has to be ingested. This may cause diarrhoea, which influences the outcome of the test. Recently, a new fluorescent label 9-fluorenylethyl chloroformate hydrazine (FMOC-hydrazine) was introduced, which reacts with reducing sugars to form stable and highly fluorescent single peak derivatives in organic medium. We applied this reagent to develop a sensitive measurement of reducing sugar probes in aqueous samples (e.g., urine). The presented method has a linear response for each sugar derivative between 1 and 1250 pmol with an R2 ranging from 0.9997 for lactulose to 0.9999 for rhamnose. The limit of detection, calculated as a signal-to-noise ratio of three, was 0.05 pmol for lactulose and 0.01 pmol for rhamnose, xylose and 3-O-methyl-D-glucose, corresponding to urine concentrations of 0.11 micromol/l for lactulose and 0.02 micromol/l for rhamnose, xylose and 3-O-methyl-D-glucose. Compared to other tests, the limit of detection is very low. This enabled a reduction in the enteral intake of the disaccharide lactulose from 6-10 g to 1.5 g, thereby minimizing the chance of introducing diarrhoea. The coefficient of variation was below 3% both in standards and urine samples. After spiking the urine with the saccharides a recovery of 102% for lactulose, 101% for rhamnose, xylose and 3-O-methyl-D-glucose was found. In order to evaluate the presented method we compared the lactulose rhamnose ratio measured in urine of healthy human volunteers and kept the ingested dose in agreement with literature values. Furthermore, the ratio was measured after 3, 6 and 9 h to establish the minimal response time required to measure correct ratios. We found that even after 3 h the ratio was stable at a value of 0.0133 which is comparable to literature values (0.008-0.052).