The connecting tubule (CNT) of the rabbit is a distinct anatomic and f
unctional segment, but little is known about its role in potassium sec
retion. We have examined potassium secretion in the CNT using the tech
nique of in vitro microperfusion. Chemical flux was determined using a
n ion-sensitive microelectrode and rubidium flux was used to estimate
the unidirectional lumen-to-bath potassium flux. In control tubules, t
he transepithelial potential difference was -4.1+/-1.1 mV and there wa
s a net potassium flux (J(K)) of -17.94+/-4. 53 pmol mm(-1) min(-1) th
at was stable for up to 3 h. The lumen-to-bath Rb-86 flux was 2.78+/-0
.38 pmol mm(-1) min(-1). This value represents 12% of the calculated b
ath-to-lumen flux of 22.32+/-2.51 pmol mm(-1) min(-1). In comparing po
tassium transport in the CNT to that of the cortical collecting tubule
, the net flux and unidirectional fluxes were significantly greater in
the CNT In the CNT ouabain 0.01 mM decreased J(K) from -25.25+/-2.81
to -2.07+/-1.94 pmol mm(-1) min(-1), a value not significantly differe
nt from zero. Potassium flux in the CNT was flow-dependent. As the per
fusion rate was increased from 5.0+/-0.5 to 13.9+/-1.1 nl/min, net J(K
) increased from -20.65+/-5.47 to -44.29+/-13.32 pmol mm(-1) min(-1).
When the perfusion rate was further increased to 26.1+/-1.9 nl/min, th
e net J(K) increased to -57.28+/-12.38 pmol mm(-1) min(-1), a value si
gnificantly greater than that obtained at a nominal perfusion rate of
5 nl/min. Both the lumen-to-bath and bath-to-lumen potassium fluxes in
creased as the perfusion rate increased. These studies are consistent
with an important role of the CNT in renal potassium excretion.