Dye transport and the bright screen

In an attempt to label embryos, we injected distal gonad arms with fluorescent dyes. To our surprise, the labeling failed, because the dyes were rapidly transported from the gonad to the intestine. We tried three dyes that differ in charge: hydroxypyrenetrisulfonic acid (HPTS), which, with three sulfonic acid groups, is always highly charged at neutral pH, carboxyfluorescein (CF), which has a single carboxylic acid group and therefore will carry a single negative charge most of the time and be neutral about 0.1% of the time, and rhodamine B (RB), which has a neutral resonance form at neutral pH. HPTS was in the gonad immediately after injection, but four hours later was entirely in the intestine. RB moved from gonad to intestine incredibly fast: by the time we injected the second gonad arm of a worm, most of the dye injected into the first arm had already moved into the intestine. CF moved from gonad to intestine at an intermediate rate. Injected dyes were rapidly eliminated when the worm was placed on a seeded plate lacking dye. In one RB-injected worm the anus was disrupted by clumsy injection technique. This worm became constipated and remained highly fluorescent, suggesting that dyes are pumped into the intestinal lumen and then defecated.

Since injection is too labor-intensive a way to label large numbers of worms, we tested whether we could label them by soaking. HPTS was not taken up by worms soaked in it, but worms soaked for 20 hours in high concentrations of RB or CF became brightly fluorescent throughout their bodies (including the gonad). Bacteria in the suspension improved the uptake of dye, presumably by stimulating pharyngeal pumping. When these labeled worms were placed on a seeded plate without dye, fluorescence rapidly moved to the intestine, and most was eliminated in an hour.

The dyes do not merely equilibrate between the gonad and the intestine - they move quantitatively from the gonad to the intestine. This suggests an active process for pumping dyes out of the gonad. The efficient elimination of HPTS, which should not be able to cross membranes passively (and didn't get in by soaking), also suggests an active process. The P-glycoproteins are an obvious candidate for the pumps. However, those P-glycoproteins that have been studied in the Plasterk lab are expressed in the intestine. Thus, although they are candidates for eliminating dyes from the intestine, they probably don't explain transport from gonad to intestine. Consistent with this, when we tested pgp and mrp mutants by soaking in CF then allowing them to feed in the absence of dye, we found that they efficiently concentrated dye in the intestine, but that the intestine remained fluorescent far longer than in wild-type.

By soaking the F₂ progeny of mutagenized worms in CF, then allowing them to feed on unlabeled bacteria and screening for worms that remained fluorescent, we have isolated two bright mutants. One of them looks like the pgp mutants: it concentrates CF in the intestine but is slow to eliminate it. The other retains dye in the gonad.

Many chemicals are ineffective on intact worms, even though they can be shown to be effective if they somehow can be gotten to their targets in the animal. For instance, 1 mM serotonin produces a maximal stimulation of pumping in a dissected pharynx, but 10 mM is required to maximally stimulate an intact worm. Jim Lewis showed many years ago that levamisol is more effective on cut worms than intact. That worms would have defenses against foreign chemicals is not surprising. We suspect that these defenses are more complex than the simple passive defense of a relatively impermeable cuticle. Active metabolism of foreign chemicals by the worm's array of cytochrome P450's, and elimination through the intestine by the mechanisms we describe here, could also serve as chemical defenses.