Fig 1.
Methacholine-stimulated (M-sweat) and CFTR-dependent (C-sweat) bubbles imaged in an oil layer on the forearm of an R117H-7T/F508del subject off and on ivacaftor.
Each image shows a mid-region (~22 mm2) of the stimulated field of sweat glands (full imaged field is ~63 mm2). Results are shown for 6 tests, A-C off ivacaftor and D-F on ivacaftor. Each pair of images shows ~ 18 M-sweat bubbles, each produced by a single, identified sweat gland. M-sweat (clear bubbles) accumulated during 10 min of methacholine stimulation. C-sweat (blue bubbles) accumulated during 30 min of stimulation with a β-adrenergic cocktail that included atropine to block M-sweat; this test directly followed the methacholine test at the same site. Blue dye particles dispersed in the oil during the C-sweat trials stained the bubbles blue to improve their visibility [17]. A freckle (light brown spot) and a hair (arrow, panel A) provided natural landmarks. A spot of ink placed on the freckle improved focusing and registration between M- and C-sweat trials. Calibration bar = 0.5 mm.
Fig 2.
Average volumes of secreted sweat as a function of stimulus and drug for R117H-7T/F508del subject S9
(A) Bar graphs showing mean ± SEM of accumulated M-sweat volumes per gland ± ivacaftor; 3 tests in each condition and in one blinded test, based on a 10 min monitoring period. (B) Bar graphs showing accumulated C-sweat volumes per gland ± ivacaftor; 3 tests in each condition and one blinded test, based on a 30 min monitoring period. Each bar graph represents mean ± SEM of 38–41 glands.
Fig 3.
Subject S9 (R117H-7T/F508del): gland-by-gland sweat secretion (±) ivacaftor.
Each point is jointly determined by the mean of M-sweat rates (x-axis) and C-sweat rates (y-axis). (A) off ivacaftor, (B) on ivacaftor; y-axis is 10X the axis of A. n = 38 glands. Five glands with the highest C-sweat rates off ivacaftor are labeled. Data were not corrected for sweat-losses; the estimated loss of C-sweat (0.023 nl·min-1·gl-1) is indicated by the dashed horizontal lines.
Table 1.
Summary data: R117H and WT subjects, all data, uncorrected.
Fig 4.
Magnitude of ivacaftor-induced increases in M and C-sweat secretion for the top 12 glands of R117H-7T/F508del subject S9 Data for the 12 glands with the largest mean C-sweat rates off ivacaftor.
(A) C-sweat rate increased markedly as a function of ivacaftor treatment, (P = 6.58E-09 two tailed paired t-test on logged data). (B) M-sweat increased slightly for the same glands (P = 0.02 two tailed t-test on logged data). Each point is an identified gland’s average secretion rate for 3 off and 3 on tests. Values were corrected for sweat volume losses by adding 0.023 nl/min to all M- and C-sweat rates (see text).
Table 2.
Summary data: R117H and WT subjects, top 12 glands; loss-corrected.
Fig 5.
Subject S11 (R117H-7T/R117H-7T).
S11 was tested once each at two different sites. (A-C) Responses off ivacaftor at site 1. (D-E) Responses on ivacaftor at site 2. (A, D) C-sweat bubbles off/on ivacaftor at 30 min time point. (B, E) C-sweat rate vs. M-sweat rate off/on ivacaftor. Each point represents a single test for a single gland and shows its M-sweat rate on the x-axis and its C-sweat rates on the y-axis. (C, F) C/M ratios vs. M-sweat rates off/ on ivacaftor. Each point represents a single test for a single gland and shows its M-sweat rate on the x-axis and its C/M sweat ratio on the y-axis. Note that y-axis scale for (F) is 5X that of (C).
Fig 6.
S10 (R117H-5T/F508del) had near-zero C-sweat responses.
(A) M-sweat rates across 3 off trials (open columns) 2 blind trials (gray columns,) and 2 open-label on ivacaftor trials (black columns). (B) The corresponding C-sweat rates for 46 glands in the ROI were zero across all trials. (C) Distribution of average M-sweat rates. M-sweat rates for each gland were averaged across 7 trials and the rate distribution plotted to illustrate that the lack of C-sweating does not result from unusually low M-sweat rates. (D) Examples of 3 glands that secreted on ivacaftor on trial 7; ‘b’ was the largest C-sweat bubble seen for this subject. Three different glands produced measureable C-sweat on trial 8 but were outside the ROI. (E) M-sweat bubbles at a different site (left arm) used on trial 6 on ivacaftor because the standard site was obstructed by an IV line. (F) Small C-sweat bubbles were observed for 4/35 glands at this site, but no corresponding testing was done off ivacaftor. Calibration = 0.5 mm for D, E, F.
Fig 7.
M-sweat (left column) and C-sweat (right column) bubbles for five control subjects and the R117H-7T/F508del subject S9.
The image for S9 is from one blinded test we had access to from his clinical trial [16] with ivacaftor. Images were cropped from the center of the field. Scale bar = 1 mm.
Fig 8.
Gland-by-gland comparisons of a control subject (WT14) with the R117H-7T subject (±) ivacaftor.
Each point is jointly determined by the mean M-sweat rate on x-axis and the mean C-sweat rate on the y-axis for a single gland across 3 trials; n = 60 glands for WT14 and n = 38 glands for R117H-7T. All data were loss-corrected by adding 0.023 nl·min-1 to the rates; this imposes a minimum rate, indicated by the dashed horizontal line, for glands with no visible sweating.
Fig 9.
Mean C/M ratios vs M-sweat rates for 5 control subjects and S9, R117H-7T (+/-) ivacaftor.
(A) Average C/M sweat rate ratios for all glands. Each symbol plots the mean M-sweat rate on the x-axis and C/M-sweat rate ratio on the y-axis for a single subject, and represents 37–60 glands and 1–3 tests per WT subject (Table 1). Filled circles: WT females; open circles: WT males. WT mean is also plotted. Data for the R117H-7T subject is the mean of 38 glands (+/-) ivacaftor across 3 trials in each condition. Ratios for the R117H subject were based on loss-corrected data. (B) Average C/M sweat rates for top 12 glands, loss corrected, and expressed as % WT average. Each symbol plots for a single subject the mean M-sweat rate on the x-axis and C/M-sweat rate ratio as a % of the WT average on the y-axis. Each symbol is the mean of the 12 glands for that subject with the highest average C-sweat rates across 1–3 tests (see Table 2). WT C/M mean is shown as dashed horizontal line. The 12 glands for S9 (R117H-7T/F508del) were those having the highest C-sweat rates off ivacaftor.
Fig 10.
Interactions between sweat rates, sweat losses and estimates of CFTR function.
(A) The % sweat lost decreases as sweat rate increases. The % of sweat that is lost is plotted on the y axis, a log scale, vs. the primary C-sweat rate on the x axis. This is for a constant loss of 0.023 nl·min-1·gl-1. The red bar shows the range of individual gland C-sweat response rates for the R117H subjects (Table 1). The horizontal blue bar shows the range of mean C-sweat responses for the 5 control subjects used in this study (Table 2). (B) Plots of the consequences of different loss estimates (x-axis) on the % WT estimate (y axis) for an observed C-sweat rate = 3% of WT rate (red circle), and various loss-corrected values and corresponding %WT estimates as black circles. The 0.023 nl/min estimate (arrow) corrected the rate to ~5% of WT. (C) Plots of the effect of different loss estimates on the fold-increase in C-sweat rate observed with ivacaftor. A loss estimate of 0.023 nl·min-1·gl-1, corresponds to a fold increase of 7 (arrow), while a loss estimate of 0.08 nl·min-1·gl-1 corresponds to a fold-increase of 3 [10].
Fig 11.
R117H-7T and R117H-5T function (+/-) ivacaftor: estimates from in vitro measurements of channel function and transcript analysis.
Each column indicates the proportion of CFTR function remaining for the conditions shown based on in vitro measurements of channel function and transcript analysis. (A) R117H-7T with reduced nPOγ to ~15% WT (column 2) produces CBAVD [25]. Pairing with a non-functional allele (column 3) predicts ~7.5% WT function—such subjects rarely present at CF clinics. Ivacaftor increases PO 2–4 fold to give 15% (dashed line) to 30% WT function. (B) R117H-5T. The 5T mutation causes a large decrease in full-length transcripts and when combined with R117H reduces function to ~1.5% (column 3) and to 0.75% when the other allele is non-functional (column 4). If ivacaftor increases Po 2–4 fold it should provide ~1.4–2.8% WT function to subjects with an R117H-5T/non-functional allele.