Urine Clearance Rates and Apparent Permeability of Gammarus Duebeni Exposed to Varying Conditions

It has been proposed that G. duebeni reduces osmotic stress by changing its permeability to water as the osmotic gradient between the haemolymph and the external medium varies (Bolt, 1983). This conclusion was based upon results obtained using THO as a radioactive marker to measure water exchange rates. However, until such changes in apparent permeability to THO can be confirmed by a different experimental approach, the possibility remains that these changes are an artefact of the THO method (Smith, 1967).

In an attempt to answer this question, Bolt (1983) demonstrated that G. duebeni exhibited large and sometimes rapid changes in apparent permeability in conditions of changing salinity whilst Chaetogammarus marinus (Leach) and Gammarus locusta (Linneaus) did not. This result partly confirms that the change in apparent permeability in G. duebeni is not due to experimental error but it still does not eliminate the possibility that unstirred layers (Nernst, 1904; Motais, Isaia, Rankin & Maetz, 1969) or other artefacts could be present in G. duebeni and not in G. locusta or C. marinus. To eliminate successfully the term ‘apparent’ from the definition of this phenomenon, permeability changes must be demonstrated by another technique.

THO experiments measure the dynamic fluxes of water into and out of the animal, and such exchanges of water in the presence of an osmotic gradient cause a bulk flow in the direction of the osmotic gradient. If this bulk flow is inwards, there must either be an increase of internal pressure or an equal flow of excreted fluid out of the animal. It is assumed that in G. duebeni the inward bulk flow is matched by an outward flow of urine. Thus, theoretically, measurement of urine flow and osmotic gradient should permit calculation of the half-time of exchange of water for the animal. This technique is only practicable if the bulk flow of water is into the animal and the animal is urinating. Thus it may only be used when the osmotic concentration of the haemolymph of the animal is greater than that of the external medium. In recent studies on the crab Holthuisana transversa, Greenaway (1981) investigated three glomerular filtration rate (GFR) markers commonly used in higher vertebrates, ^l4C polyethylene glycol 4000, ³H inulin and ⁵¹Cr EDTA, and concluded that ⁵¹Cr EDTA is likely to provide the most reliable estimate of filtration rates. For this reason, ⁵¹Cr EDTA was selected for the present study.

⁵¹Cr EDTA was obtained from the Radiochemical Centre, Amersham, Bucks. The preparation contained 3–7 mg ³¹Cr EDTA ml⁻¹ and 0·9–·4mg sodium edetate BPC ml⁻¹. The specific activity was 1–2 m Ci mg⁻¹ Cr, and to obtain the required activity per unit volume it was necessary to concentrate the material 10:1 by freeze drying.

Animals were injected through the cuticle of the lateral surface using a fine drawn-out Pasteur pipette and oral pressure. The animals were put in a counting tube, which was placed in a ‘well-type’ scintillation counter attached to a Panax Reigate counter. The salinity in the counting chamber was determined by controlling the salinity in the inflow reservoir. Water was pumped through the chamber in order to flush away the excreted urine.

A continuous signal was taken from the rate meter of the Panax Reigate counter and fed via a land line to an analogue filter (kemo) and then to an RCA COSMAC CDP 18S020 microprocessor for data logging. The microprocessor was programmed to sample repeatedly and digitize the signal from the Panax Reigate. These values were scanned over 60-s periods and stored in memory. The microprocessor was then used to calculate the urine loss as a percentage loss of body fluid per unit time using the following equations.

The calculated theoretical half-time of exchange of water (t_1/2calc) can be compared directly with the half-time of exchange as found by tritiated water (t_1/2THO) as both values refer to half-times of exchange of haemolymph volume. However, absolute values of both methods cannot be compared with confidence.

The half-time of exchange of water using THO was measured using the technique described in Bolt, Dawson, Inman & Lockwood (1980). The microprocessor system used to control the external medium is described in Lockwood et al. (1981).

G. duebeni were acclimated to a cycling salinity regime (2%oSW to 32%oSW over a 12 h 25 min time period at 15 °C for four cycles) and then injected with ⁵¹Cr EDTA. The radioactivity of the animals was monitored arid the half-time of exchange of water for G. duebeni (t_1/2calc) was calculated during those parts of the salinity cycle when the animal was hypertonic to the external medium (Fig. 1). Unfortunately, when the animal is hypotonic to the medium there is no urine flow and t _1/2 calc cannot be obtained. Even so, the pattern of change was clear and consistent (Fig. 1).

In order to compare further the tácale and the t _1/2 THO, another period of significant change in apparent permeability is needed when the animal’s heamolymph is hypertonic to the medium. In further experiments, G. duebeni were acclimated to 2% SW for 48 h and transferred to 100% SW. The salinity was then reduced to 2% SW in a linear fashion over 6 h using a second microprocessor. The sodium concentration of the haemolymph and medium and the apparent perme-ability using THO were measured. The results showed a significant change in t _1/2 THO while the animals were hypertonic to the medium (Fig. 2). This treatment was repeated with animals injected with ⁵¹Cr EDTA, with which the urine flow was measured, and t _1/2 calc found (Fig. 3). Results of a control experiment using G. duebeni acclimated to 2% SW, injected with ^5lCr EDTA and transferred to 2% SW are illustrated in Fig. 3. These show that absolute values of t _1/2 calc obtained from this technique should not be used as there is a steady drift of tincale in an animal at equilibrium with the external medium, demonstrating a change in clearance rates of the tracer as the time from injection increases.

Animals in the experimental regime showed a consistent change in t _1/2 calc, exhibiting a larger change in t _1/2 calc than the change shown by t _1/2 THO. However, if the slope of the 2% control is taken into account, this change would be reduced to a similar magnitude as that of the change in t _1/2 THO.

The technique of estimating the half-time of exchange of water using urine flow as an indicator of bulk flow involves many assumptions. This technique assumes that the excretion of ^5ICr EDTA is effectively measuring the urine flow, and that the ⁵¹Cr EDTA urine to blood (U/B) ratio is constant at unity throughout the experiment. It is also assumed that ⁵¹Cr EDTA is lost only through urination.

Reigel et al. (1974) compared radioactive markers and concluded that the loss of ⁵¹Cr EDTA via routes other than the urine was relatively small.

Lockwood & Inman (1973) have shown that the U/B ratio varies in G. duebeni from 1·05 in 100% SW to 1·15 in 2% SW. Although these variations in U/B ratio would affect the Locale, the magnitude of the observed changes in t_1/2 calc could not be explained by such small changes in the U/B ratio.

Using ⁵¹Cr EDTA to calculate t_1/2 as a complimentary method to the use of tritiated water as a direct marker strongly suggests that the ‘apparent’ permeability to water observed in G. duebeni is real. This conclusion is also supported by experiments showing different responses to tritiated water of G. duebeni, G. locusta and Chaetogammarus marinus (Bolt, 1983), where only G. duebeni exhibits radical changes in apparent permeability. It seems highly unlikely that changes in t_1/2 THO were artefacts of the experimental technique in G. duebeni but not in G. locusta and C. marinus.

It is, therefore, concluded that the present and earlier results taken together obviate the necessity for the term apparent’ when discussing permeability changes measured using THO as a marker in the amphipod crustaceans studied.

This work was supported by a Natural Environment Research Council Grant No. GT4/78/ALS/29. I thank Mr N. W. Jenkinson for technical support. I also thank Professor A. P. M. Lockwood for his helpful advice.