Effect of Ischemia on Capillary Pressure and Equivalent Pore
Radius in Capillaries of the Isolated Dog Hind Limb
By John N. Diana and M. Harold Laughlin
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
ABSTRACT
Measurements were made of filtration coefficients (Lp), area per unit path length
(A/Ax), and equivalent pore radius (rp) in the control state and following severe ischemia
(arrested blood flow) for periods of 30 minutes, 1 hour, and 3 hours. The value of A,/Ax
for diffusion of all lipid-insoluble substances was not changed after 30 minutes of ischemia,
but it was increased after 1 and 3 hours of ischemia. The value of rp calculated from the
theory of restricted diffusion yielded values of 34-35 A for both the control period and after
all three periods of ischemia. Combination of hydrodynamic data (Lp) and diffusion data
(Am/Ax) yielded values for r p of 23 A for control and ischemic periods. Measurements of
plasma-protein osmotic pressure, tissue-protein osmotic pressure, tissue hydrostatic
pressure, and capillary hydrostatic pressure supported the conclusion that extended
periods of arrested blood flow did not affect muscle capillary membrane porosity. In 5 of
16 hind limbs, there appeared to be a porosity change following 3 hours of arrested blood
flow. This change was demonstrated by a net decrease in plasma-protein osmotic pressure
and an increase in rp from 34 A to 54 A. Lp was not changed after 30 minutes but was increased after 1 hour of ischemia; the increase was associated solely with an increase in capillary surface area. After 3 hours of ischemia, the primary change in 11 of 16 hind limbs
was an increase in capillary surface area, although an increase in the size of the pores per
unit membrane area could not be rigorously excluded. In 5 hind limbs after 3 hours of ischemia, an increase in rp was the primary change and an increase in capillary surface area was
of secondary importance. The data indicate that the edema which occurs subsequent to
reperfusion of the vasculature after moderately long periods of severe ischemia results
from an increase in capillary hydrostatic pressure augmented by an increase in capillary
surface area not associated with an increase in membrane porosity. The rise in capillary
pressure for any given arterial or venous pressure involves a decrease in precapillary
resistance, but postcapillary resistance does not change for any given flow.
KEY WORDS
capillary permeability
tissue hydrostatic pressure
tissue-protein osmotic pressure
capillary surface area
• In 1963 Landis and Pappenheimer (1) summarized their review of the literature by stating:
The effects of arrested blood flow, and more
specifically, of hypoxia on capillary permeability and the filtration absorption mechanism are
still uncertain.
The present studies were made in an attempt to
clarify the effect of moderately prolonged ischemia
on those factors which control transcapillary fluid
movement. It was found that the edema which
occurred after blood flow was restored following 1-3
From the Department of Physiology and Biophysics, University of Iowa College of Medicine, Iowa City, Iowa 52242.
This study was supported by U. S. Public Health Service
Grants HL12563 and HL14388 from the National Heart and
Lung Institute and by Grant 72-G-12 from the Iowa Heart
Association.
A preliminary report has appeared in abstract form (Fed Proc
31:817, 1972).
Received August 30, 1973. Accepted for publication March
27, 1974.
Circulation Rtttmch, VoL 3S, July 1974
osmotic transient
edema
plasma-protein osmotic pressure
transcapillary fluid movement
hours of ischemia in skeletal muscle resulted primarily from a rise in capillary hydrostatic pressure
augmented by an increase in capillary surface area.
Arrest of blood flow for 1 hour did not alter effective
pore radius; after 3 hours of ischemia, pore radius
was unchanged in 11 experiments and increased in
5.
Methods
Experiments were performed on heparinized (10 mg/
kg, iv) mongrel dogs anesthetized with sodium pentobarbital (30 mg/kg, iv). The experimental procedure has
been described in detail previously (2). Briefly, the hind
limb from one dog was isolated with the aid of a cautery.
All vessels except the femoral artery and vein were
ligated; special care was taken to ensure a minimum
amount of blood leakage from the cut surface of the leg.
The femoral artery and vein were cannulated with
polyethylene tubing. The perfusion circuit consisted of
polyethylene tubing connected to a Sigmamotor pump
which delivered a pulsatile constant flow independent of
outflow pressure variations up to approximately 250 mm
Hg.
77
78
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
The femoral vein cannula was connected to rubber
tubing which led to a glass outflow J-tube. The level of
outflow pressure (femoral venous pressure) could be
varied by raising or lowering the J-tube. Both inflow and
outflow were measured continuously with a Biotronix
dual-channel electromagnetic flowmeter. The flowmeter
was calibrated for each experiment using blood as the
perfusate, a stopwatch, and a graduated cylinder.
Arterial pressure was measured from the inflow tubing
3-5 cm upstream from its junction with the femoral
artery. Venous pressure was measured in the femoral
vein by cannulating a branch of this vessel and inserting
the tip of the catheter into the vessel lumen.
The blood was oxygenated with an isolated lung
preparation from a second dog. At selected and arbitrary
intervals, the Po, and pH of the perfusing blood were
monitored using conventional O, and pH probes (Radiometer) to ensure that these parameters were in their
physiological ranges. In the control state pH varied from
7.42 to 7.59 and Po, ranged from 80 to 95 mm Hg (O,
content was calculated by assuming a value of 1.34 ml
O,/g hemoglobin for dog blood and measured hemoglobin concentrations varied from 17 to 19 ml/100
ml blood), but for any given experiment these parameters were stable for the entire period of perfusion. Blood
perfusing the limb was constantly stirred and maintained at 38°C by a reservoir-water bath arrangement.
Changes in limb weight were measured on a scale
adapted with a variable linear differential transformer.
Sensitivity could be varied but was normally set to give a
deflection of 15-20 mm for 1 g of weight change.
CAPILLARY FILTRATION
Venous pressure was elevated in increments of 10, 15,
and 20 mm Hg. For each elevation of venous pressure
there was an initial rapid gain in weight resulting from
vascular pooling of blood (3-5) and a slow gain in weight
representing filtration of fluid from plasma to tissue.
Capillary filtration coefficients (Lp) were calculated
using the slow component of weight gain as ml/min (mm
Hg pressure difference across capillary) ~\ 100 g
tissue)" 1 . Weight was converted to milliliters by dividing
grams by a specific gravity of 1.011 (mean value between
the specific gravity of water and that of plasma at 37 °C).
The tissue components of the hind limb are: 15% bone,
65% muscle, and 20% skin. For calculations of filtration
coefficients, the bone component was not included.
DIANA. LAUGHLIN
75, Abbott) (six experiments). We found no difference in
the results of these procedures and those in which blood
was left in the vessels.
At the beginning of each experiment sufficient blood
was obtained from both the experimental dog from which
the limb was amputated and from donor dogs to ensure a
complete change in blood volume in the reservoir between control determinations and determinations subsequent to ischemia. The blood from all dogs was mixed at
the very beginning of the experiment to ensure uniformity in the perfusion medium. The blood which was not
used during the period when control measurements were
made was placed in a polyethylene bottle and stored in
the refrigerator until after the ischemic period. Such
blood was then warmed to 37°C, oxygenated via the
isolated lung preparation, and placed in the perfusion
reservoir for use during the period when postischemic
measurements were made.
Following all three periods of occlusion, oxygenated
blood was perfused through the hind limb for 20-40
minutes before any experimental measurements were
made to ensure that transient changes in vascular tone
(similar to a reactive hyperemia) did not occur during
the experimental determinations. It was our experience
that arterial and venous pressures became stable approximately 15 minutes after reperfusion of the tissue began
and remained stable (as long as flow was not manually
changed) for the entire experimental period.
CAPILLARY PRESSURE
Capillary pressure (Pc) was calculated using the
relation Pc = QRU + Pv, where Q = blood flow, R,, =
postcapillary resistance, and Pv = femoral venous pressure. Q and Pu are measured variables, but Rv is an
isogravimetric quantity determined from the slope of the
line relating PU/ to Qi, where the subscript i refers to the
isogravimetric procedure. Ru was determined in the
control state and after periods of ischemia. Extrapolation
of the PU) vs. Qi line, which is linear in this preparation,
to its zero intercept on the pressure axis provided the
values for PC( which are reported in this study. Precapillary resistance (Ra) was calculated from the relation R,,
=• Pa - Pc + Q, where Pa is femoral arterial pressure. For
nonisogravimetric conditions, Pc was found as described
at the beginning of this section and Pa and Q were
measured. For isogravimetric conditions, all three parameters were measured.
ISCHEMIC PERIODS
INTERSTITIAL FLUID PRESSURE (OUYTON CAPSULE)
Ischemia was produced by completely occluding the
inflow tubing to stop blood flow to the limb. The venous
outflow tubing was also clamped shut a few seconds after
arterial occlusion. These occlusive procedures were
maintained for periods of 30 minutes, 1 hour, and 3 hours
in separate experiments; the occlusion was then released
and the limb was perfused again with oxygenated blood.
This procedure did not measure variations during hypoxic stress but rather tested for changes present in the
capillaries after the ischemic period. To eliminate the
possibility that blood "sludging" in the minute vessels
influenced the experimental results, several experiments
were performed in which blood in the limb vasculature
was flushed out within minutes after the occlusive
procedure and replaced with either 0.9% saline (three
experiments) or high-molecular weight dextran (Dextran
Circular plastic capsules having an external diameter
of 1.5 cm and a wall thickness of 1 mm were drilled with
approximately 150 holes using an 0.025-inch drill. One
larger hole was drilled (0.064 inches, o.d.) so that one
end of a polyethylene catheter (PE100) could be placed
in the center of the capsule lumen. The catheter was then
sealed to the capsule wall to prevent leakage from the
site at which it entered the capsule. That portion of the
catheter not within the capsule (approximately a
7-10-cm length of tubing) was hermetically sealed at its
external tip before implantation. This end, initially left
under the skin, was later unsealed and used to record
pressures from the capsule lumen by connecting it to a
Statham P23Db pressure transducer. The capsules were
implanted in the hind limbs of dogs; sterile methods were
used for all procedures. Of the 17 capsules that were
Circulation Raearch, VoL 35, July 1974
79
ISCHEMIA AND CAPILLARY POROSITY
implanted, 7 were located within a muscle bundle
(usually the semimembranosus) and 10 were placed in
the plane between the semimembranosus and semitendinosus muscles. Results obtained from both locations
were similar. Experiments were performed on these dogs
4-6 weeks after implantation. All dogs recovered nicely,
and at autopsy only 3 capsule sites appeared to be
inflamed and edematous. The fluid in these capsules
was distinctly different from that in the others in that it
contained hemoglobin. The data for these capsules are
not included in this report. All other capsule sites had
healed, and a description of their characteristics can be
found in a report by Guy ton et al. (6).
ence (Ca - Cr). If molecules leave the capillaries by
diffusion through aqueous channels, then Fick's law of
diffusion can be applied (Eq. 2). Assuming that the same
complex concentration difference across the capillary
membranes determines the partial osmotic pressure
which is measured by changing capillary (femoral venous) pressure, then van't Hoffs law as modified by
Staverman (11) (Eq. 3) can be applied. Combination of
these equations,
J.
= QP(Ca ~ C,),
J. = D,A x^
(1)
(2)
' A x
PLASMA-PROTEIN OSMOTIC PRESSURE
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
In vitro measurements of the osmotic pressure of
plasma proteins were made using an osmometer modified from those described by Hansen (7) and Prather et
al. (8, 9). An ultrafiltration membrane, Amicon PM30 or
UM10, was used to separate normal saline (0.9% salt
solution) from the plasma sample. A Statham P23Gb
pressure transducer was used to record pressures. The
osmometer had a response time of 2-5 minutes with
UM10 membranes and 1-2 minutes with PM30 membranes and was accurate to ±2 mm Hg of the pressures
calculated from equations outlined by Landis and Pappenheimer (1) with plasma of known albumin-globulin
concentration. Samples of blood were taken intermittently from both the perfusion reservoir and the venous
effluent from the hind limb for purposes of comparison.
Usually there was no significant difference between the
two. All plasma samples exhibited some hemoglobin
content (produced by the experimental arrangement),
but it was assumed that the component of measured osmotic pressure contributed by this molecule was small
and relatively constant per sample.
CAPILLARY PORE AREA PER UNIT PATH LENGTH
The osmotic transient method for determining area
per unit path length for lipid-insoluble substances has
been described in detail by Pappenheimer et al. (10) and
elaborated on by Landis and Pappenheimer (1). For the
studies presented in this report, lipid-insoluble molecules—urea, glucose, sucrose, raffinose, inulin, and Dextran 10 (Pharmacia)—were introduced into the blood
reservoir. The tendency to produce absorption of fluid
from tissue to plasma promoted by the osmotic effect of
these substances as they entered the capillary vasculature was counterbalanced by increasing capillary hydrostatic pressure through elevation of venous outflow
pressure. As the molecules diffused out of the capillaries,
the effective osmotic pressure within the capillaries was
reduced and the hydrostatic pressure (venous outflow
pressure) was decreased to that level which would
maintain an isogravimetric state (no net weight change).
The time course of the pressure curve (partial osmotic
pressure exerted by the introduced substances) varied as
a function of the size of the molecule and the amount
(molar concentration) of the substance in the reservoir
and was followed for 15-40 minutes. The value for
capillary hydrostatic pressure was calculated using postcapillary resistance, flow, and femoral venous pressure.
The rate at which molecules pass across the capillary
wall (J,, Eq. 1) can be determined from the product of
plasma flow ((?„) and arteriovenous concentration differCiraikaion Ratarch, Vol. 3S, July 1974
An = RTAca.,
(3)
yields a solution for the area per unit path length for
diffusion of the molecules which are being perfused
through the capillary vasculature, i.e.,
A
RTJ.a,
Ax
D.Air
(4)
where all factors on the right side of the equation are
known or measurable. Definitions of symbols are: J, =
net flux of solute (moles/sec 100 g"1 muscle), Qp =
plasma flow (ml/sec 100 g"1 muscle), Ca = arterial
concentration (moles/liter), Cv = venous concentration
(moles/liter), D, = free diffusion coefficient of molecule
(cm/sec x 10"'), A = area available for diffusion (cm1),
AC = concentration difference across the membrane
(moles/liter), Ax = path length through the membrane
(cm), AT = osmotic pressure across the membrane
(dynes/cm*), R = universal gas constant (liters atm/mole
degree), T = absolute temperature (degrees), and a. =
reflection coefficient for the specific solute and membrane.
Chemical analysis of glucose, sucrose, raffinose, and
inulin was done by the method of Harrison (12) after
hydrolysis with concentrated hydrochloric acid; colorimetry was performed on a Spectronic-100 spectrophotometer by Beckman. Urea was analyzed by the indophenol method described by Chaney and Marback (13).
Dextran 10 was analyzed by the anthrone technique
described by Wallenius (14). The absolute value for
concentrations of all substances was determined relative
to a plasma blank taken before the test substances were
added to the reservoir. We found, using the Harrison
method that the colorimetric reaction was time dependent. Because of this fact, several blanks having known
concentrations of sugar were read at periodic intervals
corresponding to the time intervals for reading the
experimental samples and the appropriate corrections
were made for changes in optical density.
REFLECTION COEFFICIENT
Eq. 3 includes a, because of established convention
and for purposes of illustration in the development of
future points to be presented in this report. In the
Results, this quantity is included in the formulas, but for
reasons which are elaborated on in the Discussion a, was
assigned a value of one unless otherwise specified.
EQUIVALENT PORE RADIUS
Calculations of equivalent pore radius were made by
successive approximation using the theory of restricted
80
DIANA, LAUGHLIN
diffusion outlined by Pappenheimer et al. (10), Renkin
(15), arid Landis and Pappenheimer (1) and described by
the equation:
thors as a derivation from Ohm's hydrodynamic
law:
(7)
where A, = pore area for solute movement, Ap = true
pore area, a = Einstein-Stokes radius of the diffusing
molecule (see Table 4 for specific radii used for the test
molecules), and r = pore radius. Equivalent pore radius
(rp) was also found by combining Aw/Ax data with filtration coefficients (Lp) as described by the equation
(6)
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
where AJ&x was found by extrapolation of the j
data to a molecular radius of 1.5 A, TJ = viscosity of the
filtrate (0.007 dyne sec/cm1), Lp = volume rate of
ultrafiltrate movement across the capillary per unit
pressure difference per 100 g of muscle (ml/sec [dyne/
cm 1 ]" 1 100 g~' muscle).
CYTOCHROME OXIOASE ACTIVITY
Prior to isolation of the limb, a control sample of
approximately 0.5 g of muscle tissue was taken from the
biceps femorous muscle. A second control sample was
taken after isolation and perfusion of the limb. At 0.5,
1.0, 2.0, 2.5, 3.0, 4.0, 5.0, and 6.0 hours after complete
occlusion of blood flow (ischemia) and at 45 minutes and
1.5 hours after reperfusion of the limb with oxygenated
blood, muscle tissue samples were also obtained. The
muscle samples were frozen in liquid Freon immediately
after removal from the limb and remained frozen until
time for analysis. Cytochrome oxidase activity was
analyzed by the method of Potter (16) as modified by
Holloszy (17). All calculations were made on a Linc-8
computer after the programs had been hand checked for
accuracy.
STATISTICAL EVALUATION
Each hind limb served as its own control, and
Student's (-test for paired observations was used for
statistical evaluation. When applicable, least-squares
regression analysis was used to find the best fit for lines
describing plotted points.
Results
This report can be conveniently separated into
two sections: the first section deals with the hemodynamic parameters which influence capillary hydrostatic pressure, and the second deals with the
filtration-absorption process and capillary porosity.
SECTION 1: PRE- AND POSTCAPILLARY RESISTANCE CHANGES AFTER
ISCHEMIA
The dependence of mean capillary pressure (Pc)
on arterial pressure (Po), venous pressure (Pu), and
precapillary (Ra) and postcapillary (/?„) resistances
to blood flow was first recognized by Pappenheimer
and Soto-Rivera (18) and expressed by these au-
The isogravimetric procedure (18, 19) provides a
method for quantifying changes in pre- and postcapillary resistance after ischemic periods; the
data are presented in Figures 1 and 2. The isogravimetric arterial pressure-flow relation was curvilinear (Fig. 1); following ischemia, for any given flow
the pressure difference between the point at which
arterial pressure was measured and the effective
midpoint of the capillaries (PC() was decreased.
More specifically, for blood flows between 20 and
80 ml/min, the percent decrease from control in
precapillary resistance was: 26.5% (range 24 to
30%), 15.4% (range 10 to 21%), and 37% (range 31 to
42%) for 30 minutes, 1 hour, and 3 hours of
ischemia, respectively. All changes were significant
(P < 0.01).
Although the mean data showed a net decrease
in precapillary resistance after 30 minutes of ischemia, there were some preparations (3 of 13) in
which Ra returned to control values shortly after
(2-5 minutes) reperfusion of the vessels with oxygenated blood was begun for this experimental
situation. This finding was not true for ischemic
periods 1 and 3 hours long.
The persistent precapillary vasodilation produced by ischemic periods was not duplicated in
the postcapillary segment. Postcapillary resistance
(Fig. 2) did not change from control after 30
minutes of ischemia (0.102 ± 0.021 [SD] VS. 0.100 ±
0.023), decreased slightly but not significantly after
1 hour of ischemia (0.104 ± 0.030 vs. 0.097 ±
0.019), and decreased significantly (P < 0.01) after
3 hours of ischemia (0.133 ± 0.019 vs. 0.092 ±
0.022). Resistance values are expressed as ml/min
mm Hg" 1 and are mean values for 13, 9, and 16
experiments for 30 minutes, 1 hour, and 3 hours of
ischemia, respectively.
Since for given arterial and venous pressures
capillary pressure depends solely on the ratio of
postcapillary resistance to precapillary resistance
as required by Eq. 7, these data suggest that Ra
decreases proportionately more than RD, which
would promote a rise in capillary hydrostatic
pressure (after reinstitution of blood flow) and
might explain, in part, the mechanism of edema
Circulation Ratarch. Vol. 35. July 1974
81
ISCHEMIA AND CAPILLARY POROSITY
15
10
30 U N ISCHEMIA
-
3 0 MIN ISCHEMIA
• CONTROL SLOPE-.102
X ISCHEMA SLOPE • .100
15
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
x 10
E
e
I HR ISCHEMIA
• CONTROL SLOPE « .104
ISCHEMIA SLOPE • .097
x
o_
15
10
- 3 HR ISCHEMIA
• CONTROL SLOPE'.134
X ISCHEMIA SLOPE > .092
3 HR ISCHEMIA
20
K>
20
30
40
50
60
70
80
Oi (ml/min)
FIGURE 1
Precapillary resistance change in isolated dog hind limb following a 30-minute period of ischemia (A), a 1-hour period of
ischemia (B), and a 3-hour period of ischemia (C). Points
plotted are mean values for 13. 10. and 16 hind limbs for A, B,
and C, respectively.
formation subsequent to a period of arrested blood
flow when the vasculature is reperfused.
The data presented in Table 1 also seem to
indicate that Ra decreases proportionately more
than Rv. The values for Q, Pa , and Pu presented in
Table 1 are mean values for eight experiments in
which inflow to the hind limb was altered in a
series of step increases or decreases and Pa and PD
Circulation Raeairh.
Vol. 35. July 1974
40
60
80
Qi (ml/min)
100
FIGURE 2
Postcapillary resistance change in isolated dog hind limb
following a 30-minute period of ischemia (top), a 1-hour period
of ischemia (middle), and a 3-hour period of ischemia (bottom).
Points plotted are mean values for 13, 10, and 16 hind limbs for
30 minutes, 1 hour, and 3 hours of ischemia, respectively.
either increased or decreased in association with
flow. Using these values for Pa and Pv and the data
presented in Figure 1 (which shows an increase in
Ra with a decrease in flow), Pc can be calculated
from Eq. 7. The last three columns in Table 1 show
that the change in Pc from control after periods of
ischemia was always an increase for any given
arterial or venous pressure. The magnitude of the
increase in Pc was small (mean value ranged from
+0.5 to +1.4 mm Hg) but significant (P < 0.001)
82
DIANA. LAUGHLIN
TABLE 1
Changes in Capillary Hydrostatic Pressure following Ischemia
Blood
flow
(ml/min)
10
20
30
40
50
60
70
80
90
P.
Pa
(mm Hg)
(mm Hg)
45
65
82
97
113
126
142
161
174
0.5
1.7
2.8
5.7
6.6
9.7
11.5
14.6
17.9
ft.
(mmHg/
[ml/min])
Ra
(mm Hg/
[ml/min])
0.102
3.84
3.21
2.83
2.56
2.34
2.14
1.97
1.82
1.77
Ischemia
ft)
Control
RJR.
Pc
(mmHg)
(mm Hg/
[ml/min])
Ischemia for 3C • Minutes
1.7
0.027
3.7
0.032
5.6
0.036
9.2
0.040
11.1
0.044
15.0
0.048
18.0
0.052
22.4
0.056
26.5
0.058
Ra
(mm Hg/
[ml/min])
P,
RJRa
(mmHg)
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
0.100
3.15
2.32
2.02
1.82
1.69
1.57
1.47
1.38
1.26
0.032
0.043
0.050
0.055
0.059
0.064
0.068
0.072
0.079
1.9
4.3
6.6
10.5
12.5
16.7
19.8
24.4
29.3
0.2
0.6
1.0
1.3
1.4
1.7
1.8
2.0
2.8
1.4
0.097
3.70
2.46
2.13
1.93
1.79
1.59
1.46
1.36
1.29
0.026
0.036
0.046
0.050
0.054
0.061
0.066
0.071
0.075
1.6
3.9
6.3
10.0
12.0
16.4
19.6
24.3
28.8
0.1
0.1
0.5
0.3
0.2
0.6
0.7
1.0
1.1
0.51
0.092
2.37
2.18
1.68
1.39
1.31
1.19
1.06
0.98
0.91
0.039
0.042
0.055
0.066
0.070
0.077
0.087
0.094
0.101
2.2
4.6
6.9
11.4
13.6
18.0
21.9
27.2
32.2
0.2
0.2
0.1
0.5
0.4
0.5
0.8
1.0
0.9
0.51
MEAN
45
65
82
97
113
126
142
161
174
0.5
1.7
2.8
5.7
6.6
9.7
11.5
14.6
17.9
0.104
4.55
3.10
2.57
2.25
2.03
1.88
1.74
1.64
1.55
hchemia for 1 Hour
1.5
0.023
0.034
3.8
5.8
0.040
9.7
0.046
11.8
0.051
15.8
0.055
18.9
0.060
23.3
0.063
27.7
0.067
45
65
82
97
113
126
142
161
174
0.5
1.7
2.8
5.7
6.6
9.7
11.5
14.6
17.9
0.134
3.70
2.96
2.51
2.24
2.02
1.85
1.69
1.55
1.42
Ischemia for 3 Hours
2.0
0.036
4.4
0.045
6.8
0.053
10.9
0.060
13.2
0.066
17.5
0.072
21.1
0.070
26.2
0.086
31.3
0.094
TAP,
(mm Hg)
J
10
20
30
40
50
60
70
80
90
MEAN
10
20
30
40
50
60
70
80
90
MEAN
8ECTION 2: EFFECTS OF ISCHEMIA ON THE FILTRATION-ABSORPTION
PROCESS
This section will present, in sequence, the
effects of ischemia on (1) the hydraulic conductivity of capillaries (Lp), (2) plasma-protein osmotic
pressure (irPI), (3) interstitial fluid pressure (PT),
(4) effective surface area per unit path length
available for fluid movement (Aw/Ax), and (5)
equivalent capillary pore radius (rp).
Effect of Ischemia on Capillary Filtration Coefficients in the Isolated Dog Hind Limb.—In the
control state, arterial and venous pressures were
adjusted so that the limb remained isogravimetric.
That is, filtration and absorption were exactly
balanced, and there was no change in weight.
Blood flows ranged between 4 and 8 ml/min 100 g~'
tissue (excluding bone). Femoral venous pressure
was then elevated in increments of 10, 15, and 20
mm Hg; the sequence for the increases in venous
pressure was random. Each separate elevation of
venous pressure was maintained for 5 minutes;
then outflow pressure was lowered so that femoral
venous pressure returned to the control value and
the limb returned to an isogravimetric state. This
procedure was followed in the control state and
after periods of ischemia of 30 minutes, 1 hour, and
3 hours when blood flow rates were approximately
the same as they were during the control period.
To produce absorption of fluid from tissue to
plasma, venous pressure was lowered below the
control value needed to maintain an isogravimetric
state.
Circulation Research. VoL 35. July 1974
83
ISCHEMIA AND CAPILLARY POROSITY
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
The rate of fluid movement was plotted against
the pressure difference across the capillary membrane (P c - PC/); the results are shown in Figure 3.
The data indicate that the capillary filtration
coefficient (Lp) did not change from control following 30 minutes of ischemia, the values for Lp being
0.0087 ± 0.0009 (SD) and 0.0086 ± 0.0007, respectively (mean values for 13 hind limbs), for this
experimental situation (Fig. 3A). Periods of ischemia of 1 hour and 3 hours significantly (P <
0.001) increased Lp . The data are presented graphically in Figure 3B and C, the values for Lpbeing
0.0086 ± 0.0009 and 0.010 ± 0.0008 for control and
1 hour of ischemia, respectively (mean values for 10
hind limbs), and 0.0092 ± 0.0006 and 0.0118 ±
0.001 for control and 3 hours of ischemia, respectively (mean values for 11 hind limbs). For reasons
©
FILTRATION
FILTRATION
ml/min/IOOg
s—\
.20.
which are discussed later, the data for 3-hour
periods of ischemia were separated into two groups.
For the second group of hind limbs (five experiments) after 3-hour periods of ischemia Lp was
almost three times greater than the control value;
the data are presented in Figure 3D.
Isogravimetric Capillary Pressures after Periods
of Ischemia.—The values for PC( (isogravimetric
capillary pressure) which were obtained in the
control state and after 30 minutes, 1 hour, and 3
hours of ischemia are presented in Table 2. There
was no significant difference between the values for
control and ischemia periods except for those for
group 2 after 3 hours of ischemia. For this group,
the value for PCj was significantly decreased (P <
0.01).
Plasma-Protein Osmotic Pressure.—The osmotic
\BJ
.15-
/
.10-
• Control
* 30 min
iichemia
Slope • .0087
.05-5
y
.15-
.10-
HO
ml/min/100 g
20-
F'X*
05-
/
5
/
-.02
15
10
5
20
10
-os
PQ-Fc.(mm Hg)
ABSORPTION
ml/min/IOOg
©-
FILTRATION
ml/min/100 g
/
.6-
/v
//
.05-
i
/
J5•Control
Slopf " 0 0 9 2
i3hr noMmia
Slopi- .0118
O5.
5
ABSORPTION
ml/min/100 g
/
O-
-5
-.05
l6
15
20 25
-19
x
• Control
Skip«'j0O90
« 3hr itchemla
Slope-.024
1/
-5
PC-PC; (mm Kg)
5
JTX
—•05
•JO
x
Xi
20-
s
/
<
/
.10-
* /
20
P C - P C (mm Hg)
FILTRATION
3 0 , ml/mln/100 g
.20
15
.rIO
-.04
ABSORPTION
ml/min/100 g
-O
• Control
Slope >. 0084
11 hr nchemia
Slope'.0111
/ /
• -V1
ABSORPTION
mlAnin/IOOg
O
O
20
Pr-Pr. (mm Hg)
^i
25
in
- v
FIGURE 3
Effect of ischemia on the capillary filtration coefficient (Lp) in the isolated dog hind limb. Lp did not
change from control after 30 minutes of ischemia (A), but increased following 1 hour (B) and 3 hours
(C) of ischemia. Points are mean values for 13, 10, and 11 hind limbs in A, B, and C, respectively.
After 3 hours of ischemia, 5 hind limbs exhibited increased porosity; Lpfor these limbs is presented in
D.
Circulation Raearch. Vol. 3S, July 1S74
84
DIANA. LAUGHLIN
TABLE 2
periods of ischemia of 30 minutes, 1 hour, and 3
hours (group 1 only). The regressions of PC/ on irPl of
arterial blood measured in vitro are expressed by
the equations:
Isogravimetric Capillary Pressures {mm ///?) after Ischemia
Group 1
IscheIschemia
mia
Con(30 Con(1
Expt. trol min) trol hour)
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
16.0
20.8
17.2
11.0
11.6
14.3
21.4
13.8
9.8
14.6
12.1
16.7
18.1
15.2
3.62
1
2
3
4
5
6
7
8
9
10
11
12
13
14.2
18.8
17.7
10.0
10.7
15.1
20.1
12.2
9.0
16.4
14.2
15.1
18.5
MEAN 14.8
± SD
3.53
Group 2
Control Pei =
IscheIschemia
mia
Con(3
Con(3
trol hours) trol hours)
12.6
14.5
10.8
17.0
12.2
15.2
12.8
13.2
16.3
12.9
13.5
16.4
9.5
19.7
11.2
15.7
12.3
12.1
14.8
13.3
16.0
19.9
13.2
12.4
14.8
16.1
13.2
11.2
15.3
9.9
12.4
14.5
17.2
12.7
13.0
14.2
16.7
12.8
10.1
14.8
10.2
12.9
14.5
18.4
17.9
17.0
20.4
9.0
8.8
12.9
13.0
14.2
13.8
1.95
135
2.92
14.0
2.77
13.6
2.26
17.6
2.15
11.6
2.50
S
30-
'/
•520,
i
E
20-
V
•
10-
A *
if
¥
C = .91 n
-(0
15
20
rtp^ (mm Hg)
/
X
i^lO-
x
p
-3.2
c1= B9np,
' '
°" 5K)
/'
_§
<
| 10-
5-
5
i
E 15-
x
X
P
^/
/
20-
X
15-
'•A
i?
C )
/
1
JI5-
(
25-
25-
-5
(10)
In Figure 4C the data are presented for five hind
limbs in which there was a significant difference
between the value of PCl found in the control state
and that found after 3 hours of ischemia (group 2).
The number of experimental points is not sufficient
to allow as rigorous an analysis as that presented in
Figure 4A and B, but it is readily apparent that
there is a significant deviation from the broken line
which represents the ideal relation which would
exist if PCl were equal to the plasma-protein osmotic pressure measured in vitro. More importantly, there is a significant difference (P < 0.001)
between these data and those found for the control
state or the other ischemic periods, which is strong
evidence supporting the fact that for these hind
limbs the permeability of the capillary wall to
protein was altered.
Interstitial Fluid Pressure Measured by Implanted Perforated Capsules.—Perforated capsules
were implanted in the hind limbs of 17 dogs, and
experiments were performed 4-6 weeks after im-
B
( ^ /)
30-
25-
/
(9)
"N
)
(
- 2.7.
Ischemia PCl = 0.89*-^ - 3.2.
pressure of the plasma proteins (*>,) was measured
in vitro as described in Methods. Data showing the
relation between Pct and irP, for 21 hind limbs and
21 separate determinations of xP, in the control
state (A) and after ischemic periods (B) are presented in Figure 4. The data in Figure 4B are for
30-
0.91TTP;
2b
/
5
-5
to
"PI
15
20
(mm Hg)
5-
5
25
-5
10
15
20
25
rip) (mm Hg)
--10
-10
FIGURE 4
Mean capillary hydrostatic pressure (Pci) required to prevent net fluid movement is slightly less than osmotic pressure produced by
plasma proteins measured in vitro by an osmometer. Broken lines represent the ideal relation, i.e., PCI - x H measured in vitro. A:
Control hind limbs. B: 30-minute, 1-hour, and 3-hour periods of ischemia. C: Five hind limbs which appeared to demonstrate loss in
plasma proteins from the intravascular space after a 3-hour period of ischemia.
Circulation Ratwrh. Vol. 35, July 1374
85
ISCHEMIA AND CAPILLARY POROSITY
plantation. Table 3 shows the mean pressures for
14 capsules obtained (1) in the control state shortly
after the limb had been isolated and perfused but
prior to any experimental manipulation, (2) during
elevation in venous pressure (&PV 23 mm Hg) in the
control state, during elevation in venous pressure
(AP0 22 mm Hg) after 30 minutes or 1 hour of
ischemia and during elevation in venous pressure
TABLE 3
Interstitial Fluid Pressure Changes as Measured by Implanted
Perforated Capsules
Intracapsular pressure
(mm Hg)
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Begin- Prening of i
elevaExpt. expt.
tion
Increase venous
pressure
Posti
1 min 3 min 5 min
- ClcVfl-
T?nA n f
EJIILI
tion
Ul
expt.*
Control (AP r •. 23.2 mm Hg)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
MEAN
+ 2.2
+ 2.0
+ 1.4
+ 1.6
-0.5
-0.5
+ 1.8
+ 1.4
+3.2
+3.2
+4.1
+3.8
+ 1.6
+ 1.2
-1.3
-1.3
-0.2
+0.2
+ 1.4
+ 1.1
+ 2.1 + 2.2
+0.5
+0.7
+ 1.0
+0.8
+0.8
+ 1.0
+ 1.25 + 1.29
After 30 Minutes or 1
1
+ 2.9
+ 2.2
2
+ 1.4
+ 1.2
3
-1.8
-0.5
4
+ 1.8 + 1.6
+4.1
5 +3.2
6
+3.9
+ 3.8
7
+ 1.6
+0.8
8
-1.0
-1.3
MEAN + 1.53 + 1.46
9
10
11
12
13
14
MEAN
After 3
-0.2
+ 1.1
+ 2.1
+0.5
+ 1.0
+0.8
+0.88
+ 2.7
+ 1.8
+ 0.0
+ 1.8
+ 3.7
+ 4.7
+ 1.4
-0.9
+ 0.7
+ 1.8
+ 2.6
+ 1.3
+ 1.3
+ 1.5
+ 1.81
+ 2.9
+ 1.9
+0.2
+ 2.1
+3.9
+5.0
+ 1.5
-0.8
+0.8
+ 2.1
+2.6
+ 1.5
+ 1.6
+ 1.6
+ 1.92
Hour of Ischemia
+3.8
+ 2.1
-1.0
-1.1
+ 1.8
+ 1.7
+4.9
+4.8
+ 4.3 +4.5
+ 2.4
+ 1.8
-0.2
-0.2
+ 2.11 +2.29
+ 3.6
+ 2.0
Hours of Ischemia (AP,
-0.4
-0.5 -0.2
+2.6
+ 1.4 + 1.9
+3.6
+ 3.1 + 3.5
+2.6
+ 1.2 + 1.9
-0.4
-0.5 -0.2
+5.5
+4.8 +5.4
+2.17 + 2.68 +2.90
+3.2
+ 2.0
+0.4
+ 2.3
+4.0
+5.4
+ 1.6
-0.6
+0.9
+ 2.4
+ 2.6
+ 1.7
+ 1.9
+ 1.7
+ 2.11
+ 2.4
+ 1.7
+0.1
+ 1.9
+3.6
+4.5
+ 1.5
-1.0
+0.6
+ 1.9
+ 2.1
+ 1.2
+ 1.3
+ 1.4
+ 1.63
!(AP. +4.0
+ 2.2
-0.9
+ 1.8
+5.2
+4.7
+ 2.5
-0.2
+2.41
218 mm Hg)
+3.2
+3.5
+ 2.0
+ 1.8
-1.2
-0.5
+ 1.7
+ 1.9
+4.6
+4.8
+4.1
+4.0
+ 2.0 + 2.3
-0.6
-0.5
+ 1.99 + 2.15
- 24.6 mm Hg)
-0.5
-1.0
+3.7 +3.7
+3.7 +3.6
+3.8 +2.6
-0.5
-1.0
+5.7 +5.0
+3.25 + 2.68
+3.1
+ 2.0
+0.0
+ 2.0
+3.7
+4.7
+ 1.5
-0.6
+0.6
+ 1.4
+ 2.2
+ 1.7
+ 1.2
+ 1.6
+ 1.79
-0.6
+3.8
+3.8
+2.8
-0.6
+5.3
+ 2.72
AP. - increase in venous pressure above that level required
to maintain the limb in an isogravimetric state.
* 3-4 hours after isolation and beginning of perfusion except
for 3-hour periods of ischemia for which the total time of the
experiment ranged from 5 to 6 hours.
Ciiruhtum Rtuarch. Vol. 35, July 1974
(APV 25 mm Hg) after 3 hours of ischemia for group
1 hind limbs only for three different time periods,
and (3) at the end of the experiment, which
corresponds to approximately 3-4 hours of isolation
and perfusion for all experimental procedures except for 3-hour periods of ischemia for which the
total time involved was 5-6 hours. Similar data to
those presented in Table 3 were obtained during
elevations in venous pressure of approximately 10
and 15 mm Hg above control, but they are not
reported since the greatest change in capsular
pressure occurred during the largest increase in
venous pressure. For control hind limbs, capsular
pressure at the very beginning of the experiment
had a mean value of +1.25 mm Hg (range -1.2 to
+ 3.8); at the end of the experiment after 3-4 hours
of isolation and perfusion (which included all
manipulations with ischemic periods, solute molecules, filtration and absorption, etc.), the mean
value was +1.79 mm Hg (range -0.6 to +4.7). It
can be seen from Table 3 that the mean values for
capsular pressure after periods of ischemia were not
significantly different from control values and the
range of pressure changes was generally the same.
During elevation in venous pressure to promote
filtration of fluid from plasma to tissue, mean
capsular pressure showed an abrupt increase immediately after elevation in venous pressure. This
increase was followed by a slow persistent rise. One
minute after venous pressure was increased the
change in mean capsular pressure (ACP) was 0.56
mm Hg, and between 1 and 5 minutes ACP was 0.29
mm Hg (mean values for all experiments). There
was no significant difference in ACP for control and
ischemic experiments.
It was unfortunate that only one hind limb which
showed a marked increase in hydraulic conductivity of the capillaries after 3 hours of ischemia
(group 2, Fig. 3D) also contained an implanted
capsule. For this one capsule the mean intracapsular pressure at the beginning of the experiment was
+ 1.6 mm Hg. During elevation in venous pressure
to promote filtration (subsequent to a 3-hour period of ischemia), the change in intracapsular
pressure was much increased compared with the
mean data for 30 minutes, 1 hour, and 3 hours
(group 1) of ischemia. The values were: preelevation +1.8 mm Hg, 1 minute +3.6 mm Hg, 3
minutes +4.5 mm Hg, and 5 minutes +5.6 mm Hg.
ACP during venous pressure elevation was thus
almost fivefold greater for this hind limb compared
with control pressures in the same hind limb or
with mean values given in the preceding paragraph
for other non-group 2 hind limbs. Correspondingly,
DIANA. LAUGHLIN
86
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
SUCROSE
RAFFINOSE
i
K>
15
20
TIME AFTER ADDITION
OF SUCROSE (mil)
TIME AFTER ADDITION
OF RAFFINOSE ImW
5
10
15
20
INULIN
25
5
10
15
20
25
TIME AFTER ADOITION
OF IMULIN Irrin)
FIQURE 6
Example of data produced by the experimental procedure. • - control, x - 30 minutes of ischemia.
Bottom sections show partial osmotic pressure developed across the capillary membrane for the
specific molecule listed. Middle sections show the net flux of the specific molecule (J, - QP[C, - C, ]).
Top sections show A./Ax for the molecule calculated from Eq. 4. Left: Sucrose. Control: 5 g was
added to 1300 ml of blood (1.3 x 70"* molesIliter). The hematocrit was 31.3%, producing a theoretical
osmotic pressure (AITHEO) of arterial plasma of 318 mm Hg. Average Qp - 0.080 ml/sec 100 g'1
muscle and skin. Ischemia: 5 g was added to 1350 ml of blood {1.6 x 10' * moles/liter). The hematocrit
was 21.5%, producing a ATTHIO - 265 mm Hg. Average Qp - 0.146 ml/sec 100 g'1 muscle and skin.
Center: Raffinose. Control: 10 g was added to 1800 ml of blood (1.6 x 10'' moles/liter). The
hematocrit was 37.3%, leading to a ATTHEO - 288 mm Hg for arterial plasma. Average Qp - 0.15
ml/sec 100g~x muscle and skin. Ischemia: 10 g was added to 1700 ml of blood (1.4 x 10-' moles/liter).
The hematocrit was 37.3%, leading to a ATTHBO - 288 mm Hg for arterial plasma. Average Qp - 0.17
ml/sec 100 g~' muscle and skin. Right: Control and ischemia: 10 g was added to 2000 ml of blood (13
x 10~' moles/liter).. The hematocrit was 38.7%. leading to a ATTHEO - 28 mm Hg for arterial plasma.
Average Qp for control - 0.117 ml/sec 100 g'' and for ischemia - 0.122 ml/sec 100 g'1 muscle and
skin. Note that only a small portion of the theoretical osmotic pressure was needed to prevent net
absorption of fluid from tissue to plasma (bottom sections): this phenomenon was not appreciably
affected after 30 minutes of ischemia except for sucrose. The net flux of molecules (middle sections)
was also not appreciably changed except for sucrose. AJ&x (top sections) remained unchanged from
control. For sucrose the net increase in Ax and the net flux balance each other, leading to no change in
A, /Ax.
after the 3-hour period of ischemia, Lp for this
preparation was approximately four times greater
than control.
Capillary Surface Area Available for Ultrafiltrate and Solute Movement.—The effective surface
area per unit path length (AjAx) for solute movement into the pericapillary spaces was determined
using six different lipid-insoluble molecules. Experiments were performed in the control state and
following periods of ischemia of 30 minutes, 1 hour,
and 3 hours. After any ischemic period, an attempt
was made to maintain all factors such as blood
flow, reservoir volume, and total concentration of
solute reaching the limb as close to control values
as was technically feasible. In practice, this uniformity was not always accomplished, but the
results did not seem to be influenced by any change
which occurred.
Ajkx did not change from control after 30minute periods of ischemia for any solute tested.
Using sucrose, raffinose, and inulin as examples, it
can be seen in Figure 5 that combination of the
data for the partial osmotic pressure developed
across the capillaries (bottom sections) with the
Cimitatim Ratarch. Vol. 35, July 1974
87
ISCHEMIA AND CAPILLARY POROSITY
not appreciably changed from control, but, after 1
hour of ischemia (Fig. 8B) and 3 hours (group 1) of
ischemia (Fig. 8C), As/Ax was measurably increased.
Of the five experiments after 3 hours of ischemia
(group 2) in which it appeared that the permeability of the capillary wall to protein was increased
(Fig. 4C), the molecules which happened to be used
to determine AjAx were urea (one experiment),
sucrose (two experiments), and raffinose (two experiments). In one of these experiments (second
sucrose experiment), when it appeared from the
experimental results that the permeability of the
capillary wall to protein had increased, we thought
it would be of interest to see if Dextran 10 would
net flux of solute molecules across the hind-limb
vasculature (middle sections) to solve for Aj&x
(top sections) by Eq. 4 showed no change from
control. In contrast to this result, it was found that
after 1 hour and 3 hours of ischemia, the relation
between osmotic pressure and net flux was such
that an increase in AjAx was found for all solute
molecules. Examples of experiments with 1-hour
and 3-hour periods of ischemia are presented in
Figures 6 and 7. The 3-hour experiments show data
from group 1 only.
A summary of the data is presented in Figure 8;
the mean data for AjAx is plotted against the
Einstein-Stokes radius of the molecular species.
After 30 minutes of ischemia (Fig. 8A), As/Ax was
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
UREA
GLUCOSE
RAFFINOSE
AVE-94 110s cm
i
m-T
1
J
f
*
" -
AVE • 36 • 10s cm
01
:___^,_J—^__^__»__-
02
- —
AVE- 5 3 . 10scm
• * - - n - ^
1
^
—
-
AVE- 2 4 . 10s cm
05
04
03
\
^
^
^
02
01
25
20
15
10
5
5
10
15
20
25
TIME AFTER ADDITION
OF UREA (min)
5
10
15
20
25
TIME AFTER ADDITION
OF GLUCOSE (min)
5
10
15 20 25
TIME AFTER ADDITION
OF RAFFINOSE (min)
FIGURE 6
Example of data produced by the experimental procedure following 1 hour of ischemia. • = control.
x = ischemia. Left: Urea. Control: 5 g was added to 1400 ml of blood (9.4 x 10'2 moles/liter). The
hematocrit was 37%, producing a ATtTHf:t, = 1824 mm Hg. Average Qp = 0.126 ml/sec 100g'' muscle
and skin. Ischemia: 5 g of urea was added to 1500 ml ofblood (9.4 x 10 ~2 moles Iliter). The hematocrit
was 39%, leading to a AWTHF.D = 1758 mm Hg for arterial plasma. Average Qp = 0.118 ml/sec 100 g''
muscle and skin. Center: Glucose. Control: 5 g was added to 1300 ml of blood (3.5 x 10'2 moles/liter).
The hematocrit was 41%. leading to a AvTHm = 672 mm Hg. Average Qp = 0.097 ml/sec 100 g''
muscle and skin. Ischemia: 5 g was added to 1300 ml of blood (3.5 x 10'2 moles/liter). The hematocrit
was 40%, leading to a AirTHr.n = 361 mm Hg. Average Qp = 0.105 ml/sec 100 g~'. Right: Raffinose.
Control: 10 g was added to 1550 ml of blood (1.9 x 10'2 moles/liter). The hematocrit was 42%, leading
to a ATTTHEO = 361 mm Hg for arterial plasma. Average Qp = 0.121 ml/sec 100g'' muscle and skin.
Ischemia: 10 g was added to 1500 ml of blood (1.9 x 10~2 moles/liter). The hematocrit was 42%.
leading to a AttTHr.r, = 373 mm Hg. Average Qp = 0.119 ml/sec 100g-' muscle and skin. Note that for
all molecules the partial osmotic pressure developed across the capillary membrane (Aw) was not
changed from control after ischemia (bottom sections); net flux (J, = Qp [C - C,,]) was increased
after ischemia, leading to an increase in AjAx (Eq. 4) above the control value.
Circulation Research. Vol. 35. July 1974
88
DIANA. LAUGHLIN
SUCROSE
INULIN
RAFFINOSE
I
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
5
10
15
20
25
TIME AFTER ADDITION
OF SUROSE (min)
5
10
15
20
25
TIME AFTER ADDITION
OF RAFFINOSE (min)
AVfc- .095 i 1 0 cm
5
10
15
20
25
TIME AFTER ADDITION
FIGURE 7
Example of experimental data following 3 hours of ischemia. • = control, x = 3 hours of ischemia
(group 1). Left: Sucrose. Control: 5 g was added to 1400 ml of blood (1.8 x 10'2 moles/liter). The
hematocrit was 45%, leading to a ATTTHEO = 346 mm Hg for arterial plasma. Average Qp = 0.077
ml/sec 100 g~' muscle and skin. Ischemia: 5 g was added to 1400 ml of blood (1.8 x 10'2 moles/liter).
The hematocrit was 45%, leading to a AVTHEO = 365 mm Hg. Average Qp = 0.077 ml/sec 100 g~ ' muscle and skin. Center: Raffinose. Control: 5 g was added to 1500 ml of blood (1.6 x 10~2 moles/liter).
The hematocrit was 35%, leading to a AITTHEO = 166 mm Hg. Average Qp = 0.79/ ml/sec 100 g'1 muscle and skin. Ischemia: 5 g was added to 1500 ml of blood (1.6 x 10~2 moles/liter). The hematocrit was
50%, leading to a ATTTHKO = 316 mm Hg for arterial plasma. Average Qp = 0.081 ml/sec 100 g~'. Right:
Inulin. Control: 10g was added to 1400 ml of blood (2.1 x 10~* moles/liter). The hematocrit was 40%.
leading to a AirTHEO = 41 mm Hg. Average Qp = 0.102 ml/sec 100 g~'. Ischemia: 10 g was added to
1300 ml of blood (2.18 x 10'' moles/liter). The hematocrit was 46%, leading to a AvTHEO = 48 mm Hg
for arterial plasma. Average Qp = 0.068 ml/sec 100 g~l muscle and skin. For this experimental situation the partial osmotic pressure developed across the capillary membrane (Ai) is not appreciably
different for control and following ischemia (bottom sections). Net flux (J, = Qp[Ca - Cv]j is increased
(middle sections), leading to an increase in AjAx (top sections) as calculated by Eq. 4.
also escape from the capillaries. As is discussed in
more detail later, Dextran 10 does not normally
escape from the capillaries of the isolated hind
limb of the dog (a = 1). Accordingly, fresh blood
was added to the reservoir and the leg was perfused
for approximately 5 minutes in an attempt to wash
out the sucrose molecules. Dextran 10 was then
added to the reservoir; an osmotic transient was
produced. The time course of the response was
followed for 45 minutes, and an analysis was made
of the arteriovenous concentration difference of
Dextran 10. The data for this single experiment are
shown in Figure 9.
A summary of the results for the five hind limbs
in which there appeared to be an increase in
capillary permeability to protein is presented in
Figure 10; the mean data show an increase over
control values in AjAx for all solute molecules.
It is important to note that, with the exception
of the one experiment reported on earlier in this
section, it was not possible to produce an osmotic
transient with Dextran 10 in either the control
state or after all three periods of ischemia. A total
of eight experiments was performed with this substance. Once capillary pressure had been increased to counterbalance the osmotic effect of
the Dextran 10 molecules, it was found that only
a slight change in weight ( + 0.2-0.6 g) occurred for
periods of perfusion up to 1 hour. Apparently the
molecular dimensions of Dextran 10 are such that
very little escaped through the capillary walls in
these isolated hind limbs. Calculation of the theoretical osmotic pressure created by Dextran 10
molecules (van't Hoff s law) yielded a mean value
for the eight experiments of 34.8 mm Hg for an
ideal semipermeable membrane. (An average
Circulation Research. Vol. 35, July 1974
89
ISCHEMIA AND CAPILLARY POROSITY
• Control
Control
5 0 min ischemia
20
w
eg
to
3hr. ischemia
20
DO
4.0
60
e.o
2.0
00
MOLECULAR RADIUS
40
6X1
80
100
(A)
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
FIGURE 8
Restricted pore area per unit path length (AJ&x - [RTo, J,]/[D. Ax]) for diffusion of solute acro3S
capillary membranes plotted as a function of the Einstein-Stokes radius of the molecular species. A:
Control us. 30 minutes of ischemia. Points are mean values for four experiments each with sucrose,
glucose, and raffinose and three experiments each with urea and inulin. B: Control vs. 1 hour of
ischemia. Points are mean values for three experiments each with urea, glucose, sucrose, and raffinose
and four experiments with inulin. C: Control vs. 3 hours of ischemia {group 1). Points are mean values
for four experiments with urea, three experiments each with glucose, sucrose, and raffinose, and two
experiments with inulin. Each hind limb served as its own control. • - control, x = ischemia.
x
Net flux
J , «Op(Co-Cv)
'sal
51
P — « / E «.I8XI0 5 cm
*
x
00,
Partial osmotic presture
developed ocro«» the
capillary membrane
10
20
30
40
TIME AFTER ADDITION OF DEXTRAN 10 (min)
FIGURE 9
Diffusion of Dextran 10 across the capillaries of one hind limb.
Since in the control state Dextran 10 would not cross the
capillary' membrane (AJ&x - 0), this experiment would appear
to demonstrate an increase in capillary porosity following a
3-hour period of ischemia.
Circulation Ratarch, Vol. 3S, July 1974
plasma volume of 860 ml to which 10 g of Dextran
10 was added was assumed, and calculations
were based on the average molecular number of
Dextran 10 specified by Pharmacia.) The mean
value of Pc calculated (Pc = QRU + Pe) after the
venous pressure was raised to counterbalance the
osmotic effect of the Dextran 10 molecules was
36.1 mm Hg. The mean osmotic pressure of nine
samples of arterial blood measured in vitro by the
osmometer was 32.9 mm Hg. The close correspondence of these values provides evidence to
support the validity of the methods used in these
studies.
Equivalent Pore Radius.—Using the theory of
restricted diffusion, curves were computed which
gave the best fit for the experimental data. For
control experiments and experiments after 30
minutes of ischemia, a pore radius (rp) of 33.7 A
gave the best fit for the experimental points (Fig.
11A). After 1 hour of ischemia, a pore radius of
35 A was computed as compared with a pore
radius of-34 A for the control period in these same
hind limbs (Fig. 11B). Following 3 hours of ischemia, the data are again separated into two groups.
Figure 11C shows that a pore radius of 34.5 A was
the best fit for the control period and that 34.8 A
best described the points after 3 hours of ischemia
(group 1). However, for the five hind limbs which
probably had a change in permeability to protein
after 3 hours of ischemia (group 2), it can be seen
DIANA. LAUGHLIN
90
Theory of Restricted Diffusion
Thtortticot curvt for port radius of
5 3 8 A drawn from theory
of rtttrlcttd
diffusion
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Sucrose
Raffinote/
Thtorttleal curvt for port rodlut of 35.4 A
drawn from thtofj of rtitricttd dlffutlon
I
I
I
L.
«O
I
I
10
8
12
14
16
18
20
22
MOLECULAR RADIUS (A*)
FIGURE 10
A./Ax plotted against the Einstein-Stokes radius of the molecular species for fiue hind limbs which
exhibited a decrease in intravascular protein osmotic pressure. AjAx increased for all molecules, and
the equivalent pore radius found by the theory of restricted diffusion indicated an increase in
porosity.
IS
18
L6
16
•
1.4
Control
3 0 mia Ischemia
L2
L4
•
Control
*
1 hr itchemto
L2
1
LO
10
OS
as
Q6
0.4
©
06
0.4
O2
©
02
s
o
8
10
MOLECULAR RADIUS
FIGURE 11
Equivalent pore radius of hind-limb capillaries found by successive approximation using the theory of
restricted diffusion:
fc-(•-9" [>-*»® •»•(?)•• = control, x - ischemia. A: Control vs. 30 minutes of ischemia. Solid line is the theoretical curve
for a pore radius of 33.7 A. B: Control us. 1 hour of ischemia. Solid line is the theoretical curve
constructed using a pore radius of 34 A for control and 35 A after ischemia. C: Control vs. 3 hours of
ischemia (group 1). Solid lines are theoretical curves for pore radii of 34.5 A (control) and 343 A
(following 3 hours of arrested blood flow).
CiraJnlinn Rwareh.
. ."W. Juhu 1Q74
91
ISCHEMIA AND CAPILLARY POROSITY
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
(Fig. 10) that a pore radius of 35.7 A provided the
best line to describe the control data and that a
pore radius of 53.8 A was computed for the experimental points after ischemia.
Another calculation of equivalent pore radius
can be made using Eq. 6 and incorporating the
value of Aw/Ax for water, which was found by extrapolation of the lines in Figures 10 and 11 to a
molecular radius of 1.5 A, and the values of Lp
determined in the same hind limbs (Fig. 3). For
purposes of comparison, it should be noted that
following 30-minute periods of ischemia there was
no change from control in Lp or Aw/Ax. For 1-hour
periods of ischemia ALP was increased 16% and
AAW/Ax was increased 16% over control values.
Similarly, ALP was increased 28% and AAW/Ax
was increased 18% after 3-hour periods of ischemia (group 1). The values calculated for rp were
23.4 A for control and 30 minutes of ischemia and
23.4 A and 23.1 A for control and 3 hours of ischemia (group 1), respectively. The most significant
change was found in those hind limbs in which
capillary permeability to protein appeared to be
altered. In these hind limbs ALP was increased
167% and Aw/Ax was increased 25% over control.
The values for rp were 23.9 A for control and 32.0 A
following the 3-hour period of ischemia. These
results are summarized in Table 4.
Reflection Coefficient.—It has been shown in
a number of studies that the reflection coefficient
(<r.) varies between 1 and 0 and can even be negative when the solute permeability exceeds that of
the solvent for a given membrane (20, 21). The
fact that a, = 1 for Dextran 10 in this preparation
(in all but one experiment) indicates that the
capillary membranes do not completely "retain"
molecules smaller than Dextran 10 and a, must
have some value less than one for such molecules.
An estimate of the reflection coefficient was determined using the approach of Vargas and Johnson
(22) and Taylor and Gaar (23), which can be briefly
summarized as follows. In the control state,
Jv = LJ^AP where the subscript j refers to plasma proteins,
etc., that are present. When solute is added to
the reservoir, this equation expands to
- Lpot As-,,
where the subscript i refers to added solute. In
the control isogravimetric state Jvc = 0 (no net
fluid movement); therefore, AP - cjA-n-j = 0.
Since AP balances out OJATTJ and is not changed
Circulation Rotorch. Vol. 35. July 1974
when solute is added, Jvexp = -LpOiA-rrt. That is,
Jv results from at ATT, and
at = —z—-— .
Jv (volume rate of flow) is measurable from the
weight loss subsequent to an isogravimetric state
produced by the net intravascular osmotic pressure
created by the molecules which are added to the
perfusing fluid; Lp is the hydraulic conductivity of
the capillaries and can be measured as shown in
Figure 3 or by the absorption rate subsequent to the
addition of Dextran 10 to the perfusing medium.
We found that there was a difference in these two
techniques (net fluid movement measured by the
latter method per unit pressure difference across
the capillary was approximately 10% less than that
measured by the former, but our limited data precluded an analysis of the phenomenon) so we arbitrarily chose to use values of Lp determined as
in Figure 3 for this calculation. ATT, is th^ osmotic
pressure exerted by the molecule at zero time.
This procedure for determining <r, requires extrapolation of the line relating weight loss (ordinate) to
zero time (abscissa). The slope of the line is then
divided by the product of Lp and the initial osmotic
pressure of the perfusate. The values for <r thus determined (13 experiments, 3 for each molecule except sucrose and raffinose for which 2 experiments
were performed) are presented in Table 4 in association with a summary of all the experimental
data. It can be seen that a, varied from 0.058 for
urea to 0.677 for inulin. The values of A,/Ax calculated from Eq. 4 using these values for a showed
that AslAx increased with increasing size of the
solute molecule. Pore radius calculations from Eq.
6 and the extrapolated value for Aw/Ax using the
a data yielded values for rp of 128 A for the control period and after all three periods of ischemia.
In contrast to this direct application of the data
to calculate pore radius, an indirect analysis can
be made which uses the definition of a presented
byDurbin(20):
1 - a = AJAL
(ID
where A, = effective pore area available for
movement of solute molecules and Aw = effective
pore area available for movement of solvent molecules. Durbin (20) showed that in artificial porous
membranes there was good correspondence between the experimental value for a determined
using Eq. 5 and that determined using Eq. 11.
Once the effective pore radius is determined, calculation of the effective pore area for water or
92
DIANA. LAUGHLIN
TABLE 4
Summary of A/Ax Data and Pore Radius Data Using a, = / and a, - some value < 1
A
Ax
RTJ.a
iD,AT
f
ontp
107100 g)
lio -
(cm x 10 ")
AJdx
Solute
radius
(cm x
Solute 10"')
o.
Ischemia
(3 hours
Con- Ischemia Ischemia Ischemia
+A
trol (30 min) (1 hour) (3 hours) porosity)
Ischeu
(3 hoi
+A
Ischemia Ischemia
(1 hour) (3 hours) porosii
Control
Ischemia
(30 min)
23±16
23±8
23±11
23 ± 11
128 ±20
125 ±22
131 ± 36
127 ± 27
a, - /
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
1.5
H,0
2.3
Urea
3.7
Glucose
4.8
Sucrose
Raffinose 5.7
9-12
Inulin
Dextran 22.2
10
1.00
1.00
1.00
1.00
1.00
1.00
1.27
1.01
0.86
0.67
0.59
0.33
0.00
1.47
1.29
1.07
0.91
0.79
0.42
0.00
1.21
0.96
0.89
0.73
0.60
0.32
0.00
a.
1.5
H,0
2.3
Urea
3.7
Glucose
4.8
Sucrose
Raffinose 5.7
9-12
Inulin
Dextran 22.2
0.058 ± 0.024
0.093 ± 0.020
0.132 ±0.019
0.252 ± 0.025
0.677 ± 0.18
1.000
0.037
0.059
0.080
0.088
0.149
0.223
0.039
0.056
0.083
0.096
0.151
0.217
- Ok
0.045
0.075
0.120
0.120
0.199
0.284
1.50
1.31
1.14
0.96
0.79
0.44
0.00
-
J.
1.59
1.43
32 ±
1.24
1.06
0.18
\
LPAT,}
0.051
0.076
0.122
0.127
0.199
0.298
10
Solute radius is the Einstein-Stokes radius of an equivalent sphere.
solute is possible using Eq. 5. A plot of three different pore radii using Renkin's (15) theory of restricted diffusion is shown in Figure 12; the values
for 1 - a were derived from a which was experimentally measured (Table 4). There is considerable scatter of the data, but most points lie between 30 and 60 A. It was found that a was not
significantly different (P > 0.50) when it was
determined in the control state or subsequent to
an ischemic period; for this reason only mean
values ± SD are presented in Table 4 and plotted
in Figure 12.
Some mention should be made of the apparent
discrepancy which occurs with these two methods
of analysis. In the first case, the experimentally
determined value of a was applied directly to Eq.
4, resulting in the curious (and not probable) situation that AjAx increases with increasing size
of the molecule. The second method of analysis
and use of Eq. 11 is a curve-fitting process that
takes the findings of Durbin (20) on artificial
membranes, which relate a to a/r (a = molecule
size, r = pore radius) and applies them to values
found in the capillaries of the dog hind limb. Two
important points should be considered. First,
there are at present no definitive studies which
give a quantitative expression of a for different
molecules in the capillaries of the dog hind limb.
The values of a determined by the "weight-loss"
method as used in this study might not be applicable and indeed might be grossly in error.
Second, application of such values to analysis by
data found in artificial membranes might simply
introduce an error in the wrong direction which
makes the data come out in more reasonable accord with recent histological evidence on capillary
pore radius (24). If o~ for the various molecules in
the capillaries of the dog hind limb ranged between
0.5 and 1.0, all three methods of analysis would
result in no quantitative differences and equivalent
pore radius values would be very similar. The
reader is directed to discussion of this topic by
Pappenheimer (25), Johnson (26), Renkin (27),
Lifson (28), and Tosteson (29) in which the considerations listed at the beginning of this paragraph are explored in greater detail.
Circulation Ratanh. Vol. 35, July 1974
93
ISCHEMIA AND CAPILLARY POROSITY
5
6 7 8
K>
MOLECULAR RADIUS (A)
FIGURE 12
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Another approximation of equivalent pore radius using I - a
derived from an experimentally determined a. Note that there is
considerable scatter in the points but that this method also
demonstrates an equiualent pore radius of 30-60 A for both the
control period and periods following ischemia of 30 minutes, 1
hour, and 3 hours duration.
It is important to emphasize, however, that the
absolute value of a would not in any manner influence the conclusion that can be made from these
studies, which is, that following 30 minutes, 1
hour, and in 11 of 16 hind limbs 3 hours of ischemia
capillary equivalent pore radius does not change.
Cytochrome Oxidase Activity.—Cytochrome
oxidase is a hemoprotein and the terminal component in the chain of respiratory carriers found
in mitochondria. Because of its position in the
respiratory chain, this enzyme is responsible for
the transfer of electrons to their final acceptor,
oxygen. There appears to be a good correlation
between the ability of muscle to perform prolonged work and its content of respiratory enzymes (17). The use of isolated, perfused organ
systems must always provoke questions which
pertain to an abnormal physiological state. A
crude attempt was made to assess the "physiological state" of the isolated, perfused hind limb
by using cytochrome oxidase activity as an index.
The type of analysis which was employed (17)
measured the rate at which cytochrome oxidase
could mobilize the mitochondrial oxidativephosphorylation system to utilize oxygen and
create energy. As shown in Figure 13, arrested
blood flow caused a decrease in cytochrome oxidase activity, but following reperfusion of the
hind limb with oxygenated blood the cytochrome
oxidase activity returned toward normal (broken
line with crosses in Fig. 13). Points presented in
Figure 13 are mean values for three experiments
Circulation Ratwxh. Vol. 35. July 1974
during the 3-hour period of ischemia normalized
to percent activity with the cytochrome oxidase
activity of a sample of muscle taken from the hind
limb of the intact dog used as 100% activity (100%
activity = 384.2 ± 46 [SD] /iliter/g min"1)- After
3 hours of ischemia (beginning of reperfusion with
blood) two experiments showed a return toward
the control value, but the tissue for the third hind
limb showed a continual decrease in cytochrome
oxidase activity. Perhaps fortuitously, the third
hind limb was also one in which there appeared
to be a change in capillary porosity.
Other parameters of interest in assessing the
physiological state of the preparation are pH and
Po, of the blood in the control state and immediately following the ischemic period. To measure
these parameters, samples of blood were taken
from the blood reservoir and from the femoral
vein 5-10 seconds after blood flow was reinstituted. The results of these determinations are
presented in Table 5. Apparently blood provides
enough buffering action to prevent a serious fall
in pH even after 3 hours of ischemia when the expected COj content would be very high.
Discussion
The results presented in this report indicated
that, in isolated dog hind-limb preparations, when
1
2
3
4
TIME AFTER AFTER ARRESTED BLOOD FLOW (Ha«)
FIGURE 13
Cytochrome oxidase activity of muscle specimens taken from
isolated dog hind-limb preparations. Activity decreased progressively with time after blood flow was arrested but increased and
reached the control level after reperfusion of the hind limb with
oxygenated blood (broken line with crosses). Data are means for
three hind limbs except after reperfusion of the limb when two
limbs showed a return to control but one limb (broken line with
solid dot) showed a continuing decrease in cytochrome oxidase
activity.
94
DIANA. LAUGHLIN
TABLE 6
Comparison of pH and Po, of Blood in the Control State and after Ischemia
Following ischemic period
Control
Venous
Po,
Period of
ischemia
(hours)
Arterial
pH
Venous
pH
Arterial
Po,
7.42
7.42
7.44
7.56
7.59
7.54
90
85
96
94
89
92
60
52
55
65
62
63
1
1
1
3
3
3
7.48
7.40
7.47
7.48
7.47
7.55
7.23
i.£
7.41
7.39
7.23
7.33
Arterial
Venous
Expt.
pH
1
2
3
4
5
6
7.50
7.48
7.45
7.52
7.58
7.52
pH*
Arterial
Po,
Venous
Po,*
92
81
90
15
22
17
6
11
7
88
93
92
* These values pertain to venous blood which flowed from the femoral vein 5-10 seconds after reinstitution of perfusion following the
period of arrested blood flow. Venous pH and Po, returned to control values within 7-10 minutes after restoration of blood flow.
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
blood flow was reinstituted after periods of arrested blood flow 30 minutes, 1 hour, and 3 hours
in duration, there was a persistent decrease in
precapillary resistance which was sustained for
at least 2-3 hours. Postcapillary resistance was
unaffected after 30 minutes and 1 hour of ischemia but was decreased after 3 hours. Following all
intervals of ischemia, the ratio of Rv to Ra was
increased, and therefore capillary hydrostatic
pressure was elevated for any given arterial and
venous pressures.
Other factors directly related to fluid equilibrium—the osmotic pressure of plasma proteins
measured in vitro, the interstitial fluid pressure,
and the protein osmotic pressure of the perivascular space (determined indirectly)—were not appreciably changed from control values in most instances after periods of ischemia for 30 minutes,
1 hour, and 3 hours. The hydraulic conductivity
(Lp) of peripheral capillary membranes was not
increased after 30 minutes but was significantly
increased after 1 hour and 3 hours of ischemia. In
all experiments at 1 hour and in 11 of 16 experiments at 3 hours, the increase in Lp was primarily
related to an increase in the area per unit path
length (A/Ax) for fluid and solute movement; capillary porosity (rp) was not altered. In 5 of 16 hind
limbs, there appeared to be a porosity change
after 3 hours of arrested blood flow. This change
was suggested by two facts: (1) TI>, measured in
vitro was considerably higher than PCj, indicating
a loss of plasma protein from the vascular space
and (2) equivalent pore radius was greater than
normal. In one experiment, passage of Dextran
10 from capillary lumen to extravascular space,
which does not occur in normal limbs, was demonstrated (Fig. 10).
Production of Edema Subsequent to Prolonged
Periods of Ischemia.—Arterial occlusive procedures in human limbs have been used for many
years by orthopedic surgeons prior to corrective
surgery. It is a common observation that subsequent to the release of occlusion the limbs become
temporarily edematous. For occlusions of up to 1
hour in all dog limbs studied and even up to 3 hours
in 11 of 16 experiments, no capillary leakage of
plasma proteins was observed. However, there
was a decrease in the ratio of precapillary resistance to postcapillary resistance and thus an increase in capillary hydrostatic pressure relative
to arterial and venous pressures. Under these
conditions capillary hydrostatic pressure will be
increased above those factors (wp, and PT) which
promote absorption, resulting in continuous net
filtration of fluid from plasma to tissue space and
thus accounting for the edema which is observed.
The increase in Lp after 1 and 3 hours of ischemia
contributes to increasing the rate of fluid loss.
In 5 of 16 preparations after 3 hours of ischemia,
there was, in addition to these changes, an increase in capillary pore size leading to a decrease
in the ability of the capillaries to retain plasma
proteins and other large molecules. The data
suggest that, in this preparation under the present
experimental conditions, 3 hours of ischemia
might be the transition point at which skeletal
muscle capillary membrane integrity is lost.
Longer periods of arrested blood flow would
probably result in a greater number of preparations showing increased capillary porosity. Such
an increase in permeability could occur because
of either destruction or lysis of the endothelial cell
membranes which make up the tubular lining of
the capillary vessel or biochemical changes in the
Circulation Rettarch, Vol. 35, July 1974
ISCHEMIA AND CAPILLARY POROSITY
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
cellular cement between the endothelial cells
which then become disorganized, thus creating
large gaps or pores between lumen and extravascular space.
It is of some significance that the data in this
report indicate that increases in capillary pressure are not substantially attenuated by decreases
in postcapillary resistance to flow after 30
minutes and 1 hour of ischemia. The observation
that postcapillary resistance does not change appreciably becomes important in evaluating the net
increase which can occur in capillary pressure. In
this preparation the method of obtaining pre- and
postcapillary resistance requires no net transcapillary fluid movement. Considering the quantitative changes in Ru/Ra, to achieve a situation in
which there was no net fluid movement subsequent
to ischemia, it is apparent from the data (Figs. 1,
2 and Table 1) that blood flow had to be reduced
below that in the control state. This finding was
true in 32 of 39 experiments. The change in blood
flow was small because the change in capillary
pressure was small (Table 1). It follows, however,
that, in the intact organism in which arterial pressure is maintained constant, the decrease in precapillary resistance would promote a capillary inflow and pressure even higher than the data in
these studies indicate; this situation would be
especially true if postcapillary resistance did not
change or was increased by intrinsic neural or
hormonal mechanisms.
The mechanism of a sustained decrease in precapillary resistance following reperfusion of the
vasculature with oxygenated blood is not clear.
Although there is abundant evidence in the literature (30-34) showing that low Po, and increased
Pco,, pH, and metabolic by-products generally
produce relaxation of vascular smooth muscle, one
would expect that such environmental changes
would be reversed when fresh, oxygenated blood
was perfused through the hind-limb vessels. That
this reversal was not the case suggests that there
might be some biochemical or molecular alteration
in the excitation-contraction coupling mechanism
of vascular smooth muscle which remains depressed after severe, prolonged hypoxia. There are
few studies which address themselves to this important problem. Honig (35) has suggested that inhibition of smooth muscle myosin adenosinetriphosphatase by inorganic phosphate and adenosine monophosphate (AMP) might participate in
this response. Detar and Bohr (36) have suggested
that a rate-limiting step in the production of highenergy intermediates needed for contraction reCuruhtion Ratanh, Vol. 35. Jufy 1974
95
suits because of the unavailability of oxygen to the
respiratory chain involved in oxidative metabolism.
The unchanging postcapillary resistance (except after 3 hours of ischemia) is difficult to explain. One might not expect venous smooth muscle
to be quantitatively as sensitive to changes in environmental pH, Po 2 , etc. as is arteriolar smooth
muscle; indeed it often responds to stimuli in a
direction opposite to that of arterial smooth muscle. However, it has been reported by Vanhoutte
and Leusen (37) that prolonged exposure of mesenteric veins to hypoxia (Po2 40 mm Hg) depresses the reactivity to electrical «timulation but
that saphenous vein strips are unaffected. The reactivity of saphenous vein strips is reduced by prolonged anoxia (POj < 1 mm Hg). We prefer to
interpret our data in the framework of recent work
by Detar and Bohr (36) who have shown that subsequent to prolonged periods of anoxia vascular
smooth muscle can exhibit augmented contractile
responses when it is subjected to stimuli such as
epinephrine. Since postcapillary
resistance
changes as determined in our studies measure both
the capillary and venous portions of the postcapillary vascular segment, it is hypothesized that,
although the capillary portion increased its crosssectional area (increased A/Ax and Lp), the
venous portion reduced its cross-sectional area
by contraction in response to reperfusion of blood.
These opposite changes resulted in an unchanging
postcapillary resistance after 30 minutes and 1
hour of ischemia. For the 3-hour period of ischemia either the veins were unable to respond to
stimuli brought about by reperfusion of blood or
the capillary cross-sectional area was increased
proportionately more than the venous cross-sectional area was reduced and net postcapillary resistance decreased.
In this respect it is also important to note that
even though precapillary resistance was reduced
after 30 minutes of ischemia, the capillary section
of the vasculature was unaffected (A/Ax and Lp
were the same as they were during the control
period). A possible explanation for this finding is
that short periods of hypoxia affect arteriolar or
metarteriolar segments which lead to nonnutritional "shunt" vessels more than they affect those
arteriolar segments which lead to nutritional
vessels in which capillary exchange occurs. This
phenomenon would lower precapillary resistance
without affecting A/Ax or Lp.
Fluid Equilibrium after Ischemia.—The increase
in capillary hydrostatic pressure which occurs in
this preparation from alterations in hemodynamic
96
DIANA. LAUGHLIN
variables after reinstitution of blood flow subsequent to ischemia constitutes only one variable
in the factors which maintain fluid equilibrium in
most tissues. The other factors are net colloid
osmotic pressure (plasma colloid osmotic pressure [TTPI] minus tissue osmotic pressure [vT]) and
tissue hydrostatic pressure (PT) in association with
the effective capillary surface area available for
fluid movement and the characteristics of the membrane as an ultrafilter (Lp). Such factors are summarized in the Starling equation which states:
±FM = LP[(PC - TTPI) - (PT - T T ) ] ,
(12)
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
where FM = net fluid movement, which can be
positive and proceed from plasma to tissue (filtration) or negative and proceed from tissue to
plasma (absorption). Under the experimental conditions used for these experiments, the tissue was
maintained in an isogravimetric state (no weight
change) and ultrafiltrate movement was in equilibrium, that is, the volume rate of flow from plasma
to tissue was exactly balanced by the volume rate
of flow from tissue to plasma. For control conditions, measured values of TTP, in vitro (Fig. 4) had
a mean of 17.9 mm Hg and those of PT had a mean
of +1.3 mm Hg (Table 3). The extrapolation of
the line relating irPi to capillary pressure (Fig.
4A) intersected the capillary pressure axis at a
negative value of 2.7 mm Hg. If protein concentration in the capillaries were reduced to zero, then
the pressure needed to maintain an isogravimetric
state (PC() would be positive by a value equal
to tissue pressure. Hence, tissue pressure acts
to increase PC( with respect to TTPI. The negative value found by extrapolation is thus best
explained by a protein concentration outside the
capillary wall sufficient to exert an osmotic effect
of 4.0 mm Hg. That is, from Eq. 12, Pc = *>, +
(PT - nT) and, when *>, = 0, Pc = P, - jr r ,
which defines the intercept. Placing these values
in Eq. 12 yields:
0 = Lp[(15.2 - 17.9) - (1.3 - 4.0)],
(13)
where the value for Pc results from knowledge of
the other three variables.
Ischemia had no appreciable effect on these
parameters. Using the results from Figure 4 and
Table 3 and the same method of analysis, the
data may be summarized as:
0 = Lp[(14.2 - 17.4) - (1.2 - 4.4)].
(14)
Evidence to support these values derives from
the fact that isogravimetric capillary pressure (PCl,
Table 2), which is the hydrostatic pressure equal
and opposite to the forces that oppose filtration,
had a mean value of 15.0 mm Hg for the control
state and 14.2 mm Hg following all ischemic
periods (except group 2, 3-hour period of ischemia).
The factors which oppose filtration are TI>, and PT,
the sum of which for these experiments was 19.2
mm Hg for the control period and 18.6 mm Hg after
ischemia. Since the hydrostatic force (Pc) will be
less than the total force (irPi + PT) by a factor equal
to the osmotic force of the tissue proteins (n-r), we
have from Eqs. 13 and 14 (TTPZ + PT) - *T = 15.2
and 14.4 mm Hg for the control and ischemic
periods, respectively. These values are very close to
the Pc, values measured experimentally. With
negligible tissue pressure, as was found in these
experiments, Pc, is roughly equal to the protein
osmotic pressure difference across the capillary
wall. That is, Pc, = (a>; - 7rT). This value should
not change with vasoconstriction or vasodilation
unless the permeability of the capillary wall to
protein is increased or protein is transported by
some mechanism (vesicular transport).
The data are incomplete, but it is of some
interest to apply this type of analysis to the five
preparations in which capillary permeability appeared to increase. The mean value for irP, measured in vitro for these preparations was 17.3 mm
Hg (Fig. 4C). The line of best fit for these few
points intersected the capillary pressure axis at
-5.1 mm Hg. The single value for PT which can be
applied in this analysis is +1.8 mm Hg. Hence, for
these preparations:
0 = Lp[(12.6 - 17.7) - (1.8 - 6.9)].
(15)
From Eq. 15, (*>, + PT) - *> = 12.6 mm Hg,
which again reflects the value for PC( (11.6 mm
Hg) found experimentally. Perhaps of greater significance is that this relation proposes an explanation for the large difference between control and
experimental Pe, values, i.e., accumulation of protein in the pericapillary space reduces the net
amount of hydrostatic pressure which must be
present in the capillaries to maintain an isogravimetric state. It is also tempting to speculate that <x
for proteins might be less than one for this experimental condition, since apparently a was less than
one for Dextran 10 and equivalent pore radius
increased to 53.8 A. However, it is not possible to
distinguish between loss of protein through pores
and vesicular transport mechanisms that might be
stimulated following hypoxic stress.
With the exception of the latter five hind limbs,
the data indicate that the permeability of the
capillary wall to protein was not appreciably afCimilation Ramrch. VoL 35. July 1974
ISCHEMIA AND CAPILLARY POROSITY
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
fected by the three periods of ischemia. These data
are in agreement with the studies of Parving (38),
who could find no increase in the transcapillary
escape rate of albumin after 3-5 hours of exposure
to hypoxia (arterial O» saturation 60-70%) in man.
Similarly, Siggaard-Anderson et al. (39) could find
no increase in capillary filtration rate in the calf of
man following inhalation of 10% O2 for 10 minutes.
Scott et al. (40) also found no change in the rate of
weight gain of dog forelimb preparations perfused
at constant flow when they were exposed to blood
with a Po, of approximately 10 mm Hg for 10-20
minutes. In the latter study, since capillary filtration coefficients were not measured and capillary
pressure changes could not be assessed, it was not
possible to provide any direct evidence for any type
of microvascular change.
A nondetectable increase in escape rate of
plasma proteins does not mean that capillary pore
dimensions did not change after ischemia but only
implies that there was not sufficient change in pore
diameter to allow bulk passage of plasma proteins
from the vascular to the interstitial space. Thus,
the increase in Lp which was seen after 1 and 3
hours of ischemia could result from an increase in
capillary porosity, an increase in the effective
capillary surface area (A/Ax) available for ultrafiltrate flow, or both. Additionally, assuming diffusion takes place through aqueous channels (pores),
an increase in A/Ax could result from (1) an
increase in the number of pores per unit membrane
area, (2) an increase in the size of the pores per unit
membrane area, (3) an increase in the number of
pores associated with an increase in membrane
area, or (4) a decrease in the diffusion path length
through the membrane. Whatever the absolute
mechanism of the change, it appears that the
duration of the ischemic period does not progressively increase the effective capillary surface area
available for fluid movement as evidenced by the
fact that AjAx increased 16% after 1 hour of
ischemia and 18% after 3 hours of ischemia (Fig. 8).
The percent increase in Aw/Ax found by diffusion
experiments is approximately 10% less than the
percent increase in Lp after 3 hours of.ischemia
(group 1). This same phenomenon was observed in
previous experiments with histamine (2) in which
the percent increase in AjAx was 40% and the
percent increase in Lp was 55% (15% higher). The
percent increase in AjAx should equal the percent
increase in Lp if all other factors remain constant.
This supposition was true after 1 hour of ischemia;
the percent change in both parameters was 16%.
Possible explanations for the inequality between
Circulation Ratardt. Vol. 35, July 1974
97
these two measurements found after 3 hours of ischemia (group 1) is that there was a small but significant change in equivalent pore radius which
was in the range of experimental error and could
not be measured. From Poiseuille's law, if ALP is
greater than A(AW/Ax)t then pore size would
change by a factor equal to the square of the ratio
of pore radii. Thus, the difference between ALP
and A(AW/Ax) after 3 hours of ischemia (group 1)
could be explained by an increase in pore radius of
4%. Similarly, the difference between ALP and
A(AW/Ax) after 3 hours of ischemia (group 2
could be explained by an increase in pore radius of
47%. For the latter case, it is worth noting that
calculation of rp by Eq. 6 shows a 39% increase and
rp determined by Eq. 5 shows a 52% increase for
this experimental condition. These considerations
strongly suggest that the primary change in the
microvasculature following 3 hours of ischemia in
group 2 preparations was an increase in the size of
the gaps or pores available for fluid movement
rather than an increase in the surface area available, although both occurred. A rigorous conclusion cannot be made, however, because another
possibility which could account for a difference between ALP and A(Au,/Ax) is, as suggested by
Yudilevich and Alvarez (41), that water moved
through nonpore regions, that is, through the endothelial cell membrane. This possibility appears
to be supported by Perl (42) whose filtration-permeability pore model for the hind limb vasculature
predicts that 15% of the solvent will go through the
cells or nonpore regions during the filtration process. The possibility that prolonged anoxia can influence the endothelial cell membrane and make it
more readily accessible to fluid movement cannot
be easily dismissed.
Although the isogravi metric and the osmotic
transient techniques were developed by Pappenheimer et al. (10, 18) in 1948 and 1951, the methods
have not been used extensively by other investigators. This observation is especially true for the
osmotic transient method. It therefore seems important to compare the results of the present
studies with those of these early investigators.
Tables 6-8 provides this information for Pci_ -KPI .
•KT, PT, Lp, A/Ax, and rp.
Specific Permeability of Muscle Capillaries following Ischemic Periods.—If the total capillary
surface area in 100 g of muscle is computed from
the data of Honig et al. (43), who found in gracilis
muscle that approximately 250 capillaries/mm2
were open for flow (10% of the total number of both
open and closed capillaries), then 1 g of muscle
98
DIANA. LAUGHLIN
TABLE 8
easily computed that for an AjAx of 1.27 x 10°
cm VlOO g and a path length of 0.5M, the total pore
area (Ap) available for ultrafiltrate flow in 100 g of
tissue in control hind limbs is 1.27 x 10s x 5 x 10"5
= 6.35 cm2. If the total surface area (Am) is 7000
cm2, then the ratio of pore area to total capillary
wall area Ap/Am = 9 x 10~4. The area increases to
7.35 cm2 ( + 16%), 7.50 cm2 ( + 18%), and 8.0 cm3
(+26%) for ischemic periods of 1 hour, 3 hours, and
3 hours with a porosity change, respectively (the
percent change from control is indicated parenthetically after the area figures). Using a path
length of 238 A, as suggested by Perl (42) in his
constricted pore model, would yield values for Ap
(for filtration of fluid) of approximately 0.3 cm2
and a ratio of Ap to Am of 4.3 x 10~\ The percent
change in pore area for the different ischemic
periods would remain the same whichever approach is used.
In defining changes in capillary membrane characteristics, it is of some use to convert hydraulic
conductivity into filtration permeability (Pf) which
has the dimensions of cm/sec. Thus,
Comparison of A/Ax Values in Control (Nonischemic) Dog or
Cat Hind Limbs
AjAx
Molecular
Substance
H,0
Urea
Glucose
Sucrose
Raffinose
Inulin
Dertran 10
MyoRlobin
D
wt
3.4
18
60
180
342
504
1.95
0.90
0.70
0.64
0.22
0.144
0.17
5500
10000
17000
a
ref. 1*
1.5 0 .55
2.3 0 .49
3.7 0 .44
4.8 0 .39
5.7 0 .34
9.2 0 .10-0. 14
ref.
10t
Present
study
1.3
0.80
0.59
0.42
0.22
22.2
19.2 0 .03
1.27
1.01
0.86
0.67
0.59
0.33
0.00
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
D = free diffusion coefficient (cm'/sec x 10*), a - approximate molecular radius (cm x 10-*), and AjAx = restricted
pore area + path length in 100 g of muscle (cm x 10').
* These values have been modified by a defined as:
WtRTLp
' ~AnVJB0
(16)
where Wt = 100 g of hind-limb tissue (excluding
bone), Vw = partial molal volume of water, Am =
total histological surface area of capillaries available for filtration (7000 cm2), and the remaining
terms have been defined previously. P, = 0.023
cm/sec for control hind limbs and after 30 minutes
of ischemia, 0.031 cm/sec after 1 and 3 hours of
ischemia, and 0.061 cm/sec after 3 hours of ischemia with a porosity change.
Similarly, the specific permeability (P.) of the
capillary membrane, which is most often expressed
in terms of flux rate per unit concentration differ-
to- - 1.
would contain 73.4 cm' of capillary surface area
(assuming the average diameter of a capillary to be
10M and the density of skeletal muscle to be 1.069 at
37°C). This value is very close to the value of 7000
cm2 surface area/100 g muscle determined by
Pappenheimer et al.(10) but much higher than
that found by Friedman (47).
If 7000 cm2 is used for the total histological
surface of the capillaries in 100 g of muscle, it is
TABLE 7
Comparison 0fPc,,1 i>,, x r , and PT in Control Hind Limbs
Pc, (mm Hg)
MEAN
±SD
±SE
i>(mmHg)
(mm Hg)
ref. 18
Present study
ref. 18
0.95 ir,, -0.56
0.91 T « - 2 . 7
8.0-32.0
15.0
16.0
Present study
17.9
2.2
ref. 18
1.7
Present study
PT (mm Hg)
ref. 18
Present study
Assumed
negligible
+ 1.22
4.0
1.1
0.54
In ref. 18, mean Pci values were not presented. Mean *>, values (measured in vitro) were not presented for normal cat and dog plasma
and x« of the perfusing medium was varied by addition of bovine serum albumin. On p485, it is stated that for one hind limb undiluted
blood Tp, - 16.0 mm Hg.
Cimilation Reuarch. Vol. 35, July 1974
99
ISCHEMIA AND CAPILLARY POROSITY
TABLE 8
Comparison of Lp and rp Values in Control Hind Limbs
RANGE
MEAN
±SD
(ml/min mm
Hg-'lOOg- 1 )
(cm x 10-')
Present
ref. 18 study
Present Present
refl*
study
study
(Eq. 5) ref. 10* (Eq. 5) (Eq.6)
34-36
4 0 - 45
0.014
0.001
0.009
0.002
24
12
23
16
* See Table 6 for value of a in ref. 1 and ref. 10.
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
ence divided by the area of the entire membrane
surface (Am), also has the dimensions of cm/sec.
D.A,
AmAx
P.=
(17)
The specific permeability for all molecules is presented in Table 9 which shows that after periods of
ischemia of 1 hour and 3 hours the specific permeability of the membrane for various lipid-insoluble
solutes has increased. Attention is directed to the
fact that the specific permeabilities listed in Table
9 derive from the AjAx data presented in Table 4
where a = 1.
In his elegant treatment of the Pappenheimer
pore puzzle, Perl (42) points out that the osmotic
transient method determines PJa and that to
obtain true P, values the data from the osmotic
transient experiments have to be modified. Table 9
also presents P, values modified by multiplying
PJa values by the experimentally determined
value of a (Table 4). The second set of permeability
values agrees more closely with permeability values
for muscle capillaries found by tracer experiments
(45-47), and yet both the osmotic transient technique using a = 1 and tracer studies provide similar
measures of equivalent pore radius. The paradox
resolves to the value of a which obtains in hindlimb capillaries during the osmotic transient procedure. A general discussion of this problem has been
presented by Pappenheimer (48), Lifson (28), Tosteson (29), Perl (42), and Diana et al. (2) and will
not be pursued further in this report. All values
for A/Ax presented in this report vary directly with
the numerical value of a. Perl et al. (49) have
recently reported that the reflection coefficient for
NaCl in dog lung is approximately 0.3. This high
value suggests that the reflection coefficient for a
small molecule in muscle capillaries is much higher
than that predicted by the weight-loss method used
in this study. It also suggests that the magnitude of
change in A/Ax values found in this study would be
considerably decreased if the true value of a were
known.
It should be recognized that the osmotic transient technique using small lipid-insoluble substances measures the small-pore system of the
microvasculature and not the large-pore system
(50). If changes in the large-pore system (250 A
pore radius) occurred with ischemia (or with histamine [2]) this technique would not demonstrate
TABLE 9
Specific Permeability of Hind-Limb Capillaries to Various Molecules in the Control State and following Ischemia
MolecD
(cm'/sec
ular
Molecule wt
x 10*
3.40
18
Urea
1.95
60
0.90
Glucose
180
Sucrose
0.70
342
0.64
Raffinose
504
Inulin
5,500 0.21-O.26
Dextran 10,000
0.144
10
H,0
Control
61.7
28.1
11.1
6.7
5.4
1.2
0
Specific permeability = P,
P.
(cm/sec x 10")
Ischemia Ischemia Ischemia
(30min) (lhour) (3 hours)
58.8
26.7
11.4
7.3
5.5
1.2
0
71.4
35.9
13.8
9.1
7.2
1.6
0
72.9
36.5
14.7
9.6
7.2
1.6
0
DA - (moles/sec cm1 membrane per
of muscle was taken as 7000 cm*.
Circulation Ratorch. Vol. 35. July 197*
i
PJa
(cm/sec x 10*)
Ischemia
(3 hours
+A
porosity)
77.2
39.4
12.4
9.7
0.54
Ischemia
(3 hours
+A
porosity)
a
Control
Ischemia
(30min)
Ischemia
(lhour)
Ischemia
(3 hours)
0.058
0.093
0.132
0.252
0.677
0.900
1.63
1.03
0.88
1.36
0.83
0
1.55
1.06
0.96
1.38
0.79
0
2.08
1.28
1.20
1.82
1.05
0
2.12
1.36
1.27
1.82
1.10
2.28
0
0.49
1.64
2.43
mole/ml concentration difference). Total membrane surface in 100 g
DIANA. LAUGHLIN
100
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
such an occurrence. The fact that in only one hind
limb was it possible to show that Dextran 10
diffused across the capillary bed following a 3-hour
period of ischemia would suggest that the smallpore system for this hind limb had changed.
However, if there is a distribution of pore sizes from
40 to 250 A radius, some of the intermediate sized
pores might have changed; this alternative cannot
be eliminated. Of some interest is the fact that the
tail-end of the restricted diffusion curve (small
A/Ax vs. molecular radius) appears to be influenced to a larger extent than the upper portion
of the curve (large A/Ax vs. molecular radius) when
a porosity change is observed (Fig. 10). From a
purely empirical point of view, this finding is what
one would expect from a porosity change, since a
molecule the size of urea has less trouble diffusing
through small pores initially although a molecule
the size of inulin has a difficult time. The percent
change in the amount of inulin which passes by
diffusion would greatly increase if the pore size
became larger, but the diffusion of urea would be
less influenced. This concept is expressed more
rigorously by the restricted diffusion theory (Eq. 5).
Using a pore radius change from 35 to 54 A and the
Einstein-Stokes radius of the molecular species, it
can be shown that AJAP would increase 10.7% for a
molecule the size of urea and 66.3% for a molecule
the size of inulin.
Acknowledgment
The authors gratefully acknowledge the capable technical
assistance of Mrs. Beth Berg and Mrs. Huei-Li Wang.
References
recording of plasma colloid osmotic pressure of whole
blood. J Appl Physiol 24:603-605, 1968
9. PRATHER JW, BROWN WH, ZWEIFACHBW: Improved osmom-
eter for measurement of plasma colloid osmotic pressure.
Microvasc Res 4:300-305, 1972
10. PAPPENHEIMER JR, RENKIN EM, BORRERO LM: Filtration
11.
12.
13.
14.
15.
16.
17.
18.
19. DIANA JN, KAISER RS: Pre- and postcapillary resistance
during histamine infusion in isolated dog hindlimb. Am J
Physiol 218:132-142, 1970
20. DURBIN RP: Osmotic flow of water across permeable cellulose membranes. J Gen Physiol 44:315-326, 1960
21. SCHULER KE, DAMES C, LAIDLER K: Kinetics of membrane
22.
1. LANDIS EM, PAPPENHEIMER JR: Exchange of substances
through capillary walls. In Handbook of Physiology, sec 2,
vol 2, Circulation, edited by VVF Hamilton and P Dow.
Washington, D.C., American Physiological Society, 1963,
pp 961-1034
23.
24.
2. DIANA JN, LONG SC, YAO H: Effect of histamine on
equivalent pore radius in capillaries of isolated dog
hindlimb. Microvas Res 4:413-437, 1972
3. DIANA JN, COLANTINO R, HADDY FJ: Transcapillary fluid
movement during vasopressin and bradykinin infusion.
Am J Physiol 212:456-465, 1967
4. DIANA JN, SHADUR C: Effect of arterial and venous pressure
on capillary pressure and vascular volume. Am J Physiol
225:637-650, 1973
5. JOHNSON PC, HANSON K: Effect of arterial pressure on
arterial and venous resistance of the intestine. J Appl
Physiol 17:503-508. 1962
diffusion and molecular sieving through peripheral capillary membranes. Am J Physiol 167:13-46, 1951
STAVERMAN AJ: Theory of measurement of osmotic pressure.
Rec Trav Chim 70:344-352, 1951
HARRISON HE: Modification of the diphenylamine method
for determination of insulin. Proc Soc Exp Biol Med
49:111-114, 1942
CHANEYAL, MARBACH EP: Modified reagents for determination of urea and ammonia. Clin Chem 8:130-132, 1962
WALLENIUS G: Some procedures for dextran estimation in
various body fluids. Acta Soc Med Upsalien 59:69-77,
1953
RENKIN EM: Filtration, diffusion and molecular sieving
through porous cellulose membranes. J Gen Physiol
38:225-243, 1954
POTTER VR: In Manometric Techniques, edited by WW
Umbreit, RH Burris, and JR Stauffer. Burgess Publishing
Company, 1964, pp 162-164
HOLLOSZY JO: Biochemical adaptations in muscle: Effects of
exercise on mitochondrial oxygen uptake and respiratory
enzyme activity in skeletal muscle. J Biol Chem
242:2278-2286, 1967
PAPPENHEIMER JR, SOTO-RIVERA A: Effective osmotic pressure of the plasma proteins and other quantities associated with capillary circulation in the hindlimbs of cats
and dogs. Am J Physiol 152:471-491, 1948
25.
26.
27.
6. GUYTON AC, GRANGER HJ, TAYLOR AE: Interstitial fluid
pressure. Physiol Rev 51:527-563, 1971
7. HANSEN A: A self recording electronic osmometer for quick
direct measurement of colloid osmotic pressure in small
samples. Acta Physiol Scand 53:197-213, 1961
28.
8. PRATHER JW, GAAR K, GUYTON AC: Direct continuous.
29.
processes: III. Diffusion of various nonelectrolytes through
collodian membranes. J Chem Phys 17:860-865, 1949
VARGAS F, JOHNSON J: Estimation of reflection coefficients
for rabbit heart capillaries. J Gen Physiol 47:667-677,
1964
TAYLOR AE, GAAR K JR: Estimation of equivalent pore radii
of pulmonary capillary and alveolar membranes. Am J
Physiol 218:1133-1140, 1970
KARNOVSKY MS: Morphology of capillaries with special
reference to muscle capillaries. In Capillary Permeability,
Alfred Benzon Symposium II, edited by C Crone and N
Lassen. New York, Academic Press. 1970, pp 341-350
PAPPENHEIMER JR: Discussion. In Capillary Permeability,
Alfred Benzon Symposium II, edited by C Crone and N
Lassen. New York, Academic Press. 1970, pp 293-295
JOHNSON JA: Reflection coefficients of non-electrolytes in
the myocardium. In Capillary Permeability, Alfred Benzon Symposium II, edited by C Crone and N Lassen. New
York, Academic Press, 1970. pp 291-292 and 295-299
RENKIN EM: Discussion. In Capillary' Permeability, Alfred
Benzon Symposium II, edited by C Crone and X Lassen.
New York, Academic Press, 1970, pp 287-288
LIFSON N: Revised equations for the osmotic transient
method. In Capillary Permeability, edited by C Crone
and N Lassen. New York. Academic Press, 1970, pp
302-305
TOSTESON DC: Discussion. In Capillary Permeability, edited
Circulation Rtttarch. Vol. 35. July 1974
ISCHEMIA AND CAPILLARY POROSITY
by C Crone and N Lassen. New York, Academic Press,
1970, pp 658-664
30. DAUGHERTY RM JR, SCOTT JB, DABNEYJM, HADDYFJ: Local
effects of O, and CO, on limb, renal and coronary vascular
resistance. Am J Physiol 213:1102-1110, 1967
31. KONTOS HA, RICHARDSON DW, RAPER AJ, PATTERSON JL JR;
Contribution of hypercapnia and hypoxia to the vascular
response to ischemia. Clin Sci 39:203-222, 1970
32. GUYTON AC, Ross JM, CARRIER O JR, WALKER JR: Evidence
33.
34.
35.
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
36.
37.
38.
for tissue oxygen demand as the major factor causing
autoregulation. Circ Res 15(suppl I):I-60-68, 1964
DETAR R, BOHR D: Oxygen and vascular smooth muscle
contraction. Am J Physiol 214:241-244, 1968
DUUNG BR, BERNE RM: Oxygen and the local regulation of
blood flow: Possible significance of longitudinal gradients
in arterial blood oxygen tension. Circ Res 28(suppl
I):I-65-69, 1971
HONIG CR: Control of smooth muscle actomyosin by phosphate and 5'-AMP: Possible role in metabolic autoregulation. Microvasc Res 1:133-146, 1968
DETAR R, BOHR D: Contractile responses of isolated vascular
smooth muscle during prolonged eiposure to anoxia. Am
J Physiol 222:1269-1277, 1972
VANHOUTTE P, LEUSEN I: Effect of hypoxia on reactivity of
cutaneous and mesenteric veins of the dog. In Vascular
Smooth Muscle, edited by E Betz. New York,
Springer-Verlag, 1972, pp 138-139
PARVING HH: Effect of hypoxia and carbon monoxide
exposure on plasma volume and capillary permeability to
albumin. Scand J Clin Lab Invest 30:49-56, 1972
39. SIGGAARD-ANDERSON J, PETERSEN FB, KYELDSEN K: Volume
changes of the calf during ten minutes' venous stasis. Acta
Med Scand 184:289-292, 1968
40. SCOTT JB, DAUGHERTY RM JR, HADDY FJ: Effect of severe
local hypoxemia on transcapillary water movement in dog
Circulation Ratarch. Vol. 35, July 1974
101
forelimb. Am J Physiol 212:847-851, 1967
41. YUDILEVICH DL, ALVAREZ OA: Water, sodium and thiourea
transcapillary diffusion in the dog heart. Am J Physiol
213:308-314, 1967
42. PERL W: Modified filtration-permeability model of transcapillary' transport: A solution of the Pappenheimer pore
puzzle? Microvasc Res 3:233-251, 1971
43. HONIG CR, FRIERSON JL, PATTERSON JL: Comparison of
neural controls of resistance and capillary density in
resting muscle. Am J Physiol 218:937-942. 1970
44. FRIEDMAN JJ: Total, non-nutritional and nutritional blood
volume in isolated dog hindlimb. Am J Physiol
210:151-156, 1966
45. LASSEN NA, TRAP-JENSEN J: Estimation of the fraction of the
inter-endothelial slit which must be open in order to
account for the observed transcapillary exchange of small
hydrophilic molecules in skeletal muscle in man. In
Capillary Permeability, edited by C Crone and N Lassen.
New York, Academic Press, 1970, pp 647-653
46. ALVAREZ OA, YUDILEVICH DL: Heart capillary' permeability
to lipid-insoluble molecules. J Physiol (Lond) 202:45-58,
1969
47. DURAN W: Early capillary extraction of sodium in resting
and exercising canine skeletal muscle (abstr). Physiologist 16:302, 1973
48. PAPPENHEIMER JR: Osmotic reflection coefficients in capillary membranes. In Capillary Permeability, edited by C
Crone and N Lassen. New York, Academic Press, 1970, pp
278-286
49. PERL W, CHOWDHURY P, CHINARD FP: Osmotic reflection
coefficient of dog lung endothelium to sodium chloride,
glucose, sucrose, raffinose and albumin. Microvasc Res
6:125-126, 1973
50. GROTTE G: Passage of dextran molecules across the bloodlymph barrier. Acta Chir Scand [Suppl] 211:1-84, 1956
Effect of Ischemia on Capillary Pressure and Equivalent Pore Radius in Capillaries of the
Isolated Dog Hind Limb
JOHN N. DIANA and M. HAROLD LAUGHLIN
Downloaded from http://circres.ahajournals.org/ by guest on June 18, 2017
Circ Res. 1974;35:77-101
doi: 10.1161/01.RES.35.1.77
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1974 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/35/1/77
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/