Journal of Neuroscience Methods, 39 ( 1991 ) 245- 251 ~ 1991 Elsevier Science Publishers B.V. All rights reserved 0165-0270/91/$03.50
245
NSM 111292
Measurement of brain tissue specific gravity using pycnometry G e n e R o b e r t D i R e s t a ~, J o n g b i n L e e 2 and E h u d A r b i t 3 I Nuclear Medicine Research Laboratory, : Neurosurgical Research Laboratoo', and ~ Department Of Surgery, Memorial Sloan-Kettering Cancer Center, New York, NY 10021 (U.S.A.) (Received 10 June 1991) (Accepted 23 July 19911
Key words: Brain density; Water content; Specific gravity; Pycnometry In this paper we introduce and characterize pycnometry, a method used to measure fluid density, fl)r determining a tissue's specific gravity. It uses a 2-ml glass pycnometer filled with distilled water to determine a tissue sample's displacement volume. The tissue's density is determined when it's weight is divided by this volume and specific gravity is computed by dividing the tissue density by the density of water. Pycnometry was validated using pre-calibrated glass, specific gravity standards over the range 1.03-1.26, and compared to the density gradient method using rat brain tissue. We observed that the specific gravity values obtained using pycnometry were highly correlated with the specific gravity standards ( s l o p e - 1.0107, r = I).996) and with the density gradient column when tissue volumes larger than 0.120 ml were used with the pycnometer (slope = 1.0707, r - 0.9826). Good correlation was also observed between percent water content values computed using the Nelson equation with pycnometry or density gradient specific gravity values versus the measured percent water content values obtained with the wet w e i g h t / d r y weight method. Pycnometry is an accurate, reproducible technique to measure tissue specific gravity and brain edema and is best suited for use in a laboratory that engages sporadically in brain edema measurement.
Introduction
Cerebral edema, i.e., increase in brain water content, accompanies a wide variety of pathological processes in the brain and is responsible for the morbidity associated with these conditions. Cerebral edema is particularly prominent in head trauma, stroke, brain tumors, infections and metabolic encephalopathies. The measurement of edema is, therefore, an important indicator used in research in these areas. Edema has traditionally been determined using the wet w e i g h t / d r y weight technique. This method provides a measure of the percent of water content in tissue and is simple to perform
Correspondence: Ehud Arbit, M.D., Memorial SIoan-Kenering Cancer Center, 1275 York Avenue, New York, NY 10021, U.S.A.
(Hossmann et al., 1980; Hiroyuki et al., 1987). Its shortcomings include requiring relatively large samples and long drying times, typically 24 h or more, to ensure reasonable accuracy (Marmarou et al., 1982). Specific gravity measurements provide an alternative index of edema. The technique of using flotation in fluids of known density to determine specific gravity for the determination of serum protein concentrations was first reported by Lowry and Hunter (1945). Laborite and Weber (1965) employed a series of copper sulfate solutions of defined specific gravity to measure changes in edematous rat brains. Nelson et al. (1971) used a gradient column of kerosene and bromobenzene to measure changes in brain water content. The density gradient method was validated by Marmarou et al. (1978) and has since gained wide
246
acceptance among neuroscientists (Fujimoto et al., 1976; Ferszt et al., 1978; Klatzo et al., 1980; Marmarou et al., 1980; Ishige et al., 1987; Karlik and Noseworthy, 1987; Hatashita and Hoff; 1988; Nagao et al., 1988) mainly because of its simplicity and reproducibility, particularly with small samples as compared with the wet weight/dry weight method. The measure of specific gravity lends itself to conversion to percent water content with the Nelson equation (Marmarou et al., 1978; Marmarou et al., 1982). While the density gradient method is an invaluable tool for the study of brain edema, it suffers from several disadvantages. The accuracy of the method depends upon such factors as the linearity of the gradient column, condition of the calibrating standards, temperature and humidity of the dissection area, optimal sample size, and the interval between obtaining the sample and introduction into the density column. The density gradient method uses kerosene and bromobenzene, two volatile, flammable and toxic chemicals, and the column preparation is a time-tedious process. In addition consideration must be given to the tissue's lipid content in relation to its solubility in the gradient column's solutions. Lastly, errors are introduced into the measuremcnt of brain tissue water when the edema fluid contains protein (Steward-Wallace, 1939; Shigeno et al., 1982; Bothe and Hossman, 1984). In this paper we introduce pycnometry, a simple method frequently used to determine fluid density, to measure tissue specific gravity (Kitagawa et al., 1981). We evaluate its accuracy against glass specific gravity standards, determine the optimal specimen size using a uniform density material, and correlate it against both the column density gradient and wet weight/dry weight methods using rat brain tissue specimens.
Materials and methods
Pycnometer method The pycnometer method requires a micro-analytical balance, i.e., 0.01 mg resolution, and a small 2-ml glass pycnometer (Fisher Inc.) shown in Fig. 1. Its center hole was plugged with water-
PYCNOMETER
-~'~/~mar~ks ~...
Alignment ~ - ~ ' ~
Epoxy ~ugs
FLASK (2 ml) Fig. I. Standard 2-ml glass pycnometer (Fisher Scientific). Black marks placed on exterior of flask and on ground glass stopper; waterproof epoxy used to seal both stopper ends.
proof epoxy to minimize evaporative losses. Two black marks were placed on the ground-glass stopper and on the outside neck of the flask to ensure equivolume between measurement. The pycnometer was carefully cleaned prior to use and wiped dry using lint-free paper. The fluid used in our studies was distilled water which had been filtered and degassed by placing it under vacuum for 1 h prior to use. The method is used as follows with all measurements made at thc same temperature. (a) The pycnometer is filled with pure distilled water which has reached the desired temperature, i.e., 23 ° C. The stopper is introduced and its mark is aligned with the pycnometer mark. The outside of the flask is thoroughly dried with lintfree wipes and weighed (mot.). Air bubbles, if present, are eliminated by removing the stopper and immersing the flask into an ultrasonic cleaner for several seconds. (b) The tissue sample(s) is weighed. (c) Some water from the pycnometer is removed and the tissue transferred into the flask. The pycnometer is refilled with water, bubbles are removed, marks aligned and the outside of the flask is dried thoroughly. The pycnometer containing the sample is then weighed ( m r ) . (d) The tissue specific gravity is calculated by dividing the tissue density (derived from Eqn. t below) by the density of water at the measurement temperature (d,,).
247
Eqn. 1 was used to determine tissue density from the weight measurements (its derivation is presented in the Appendix):
d~=
S
(m~j-m~ +s)
xdw
were re-weighed. The difference between the tare weight and the new weight was the tissue's 'dry weight'. The percent water content is determined from Eqn. 2 below:
(1)
wet weight-dry weight % water content =
where d~ = density of the sample (g/ml), d,, = density of water ( g / m l ) at temperatures of measurement, s = weight of sample (g), mpj,= weight of pycnomcter and water (g), m t = weight of pycnometcr, water and sample (g). Now specific gravity is related to density by the relationship: specific gravity = d J d w.
Density gradient column method The density gradient single column system, Model DC-1, Techne Inc., was used in this study. The density gradient was prepared with kerosene and bromobenzene according to the method presented by Marmarou et al. (1982). The column of liquid was allowed to reach a temperature of 2 3 ° C and then calibrated using glass specific gravity standards (Techne Corporation) over the range 1.03-1.26. With a properly prepared gradient column, the specific gravity varies linearly according to the height of the column. The standards' positions within the column versus their specific gravity was plotted and evaluated for its linearity using statistical regression analysis. The column was accepted if the correlation coefficient was at least 0.9995; if not, the fluids were discarded and the gradient re-established. Specimens weighing 60 mg were lowered into the column at 5-min intervals and allowed to descend. The specific gravity of the sample was determincd from the depth of the sample read at 5 rain post-immersion. This position was related to the sample's specific gravity with the calibration plot. Wet weight/dry weight method Aluminum specimen dishes were labeled and tared prior to dissection of brain and again immediately after receiving its sample. The difference between these 2 weights is the tissue's 'wet weight'. The dishes were then placed into a drying oven adjusted to 9 0 ° C . After 5 days, they
wet weight
x 100
(2)
Specific grat,ity to percent water content com,ersion method Specific gravity values measured by either the density gradient or pycnometry were converted into percent water content using Nelson's method (Nelson et al., 1971) This technique relates spccific gravity to its approximate (g:g) H 2 0 / t i s s u e value. Surgical procedures and tissue preparations S p r a g u e - D a w l e y male rates (300-400 g) were anesthetized with a mixture of 5 m g / k g xylazine and 80 m g / k g ketamine and then killed by an overdose injection of saturated KCI. The rat brain was then quickly removed from the cranium. Large tissue samples were taken from the frontal, parietal, occipital cortex, cerebellum and medulla. Each sample was divided into two pieces, one weighing approximately 120 mg for the pycnometer or the w e t / d r y method and the second weighing approximately 60 mg for the density gradient method. Samples were stored in pre-chillcd petri dishes until used. Uniform density tissue preparation The homogenized tissue from a whole rat brain was mixed with 10 ml of warm, sterile 30% gelatine solution. The mixture was poured into a sterile petri dish, covered, refrigerated and allowed to set. Its density was determined using a 60 mg sample with the column density gradient method. The remaining matrix was cut into samples of varying sizes. The density of these samples was determined using the pycnometer method to determine the influence of sample volume on pycnometry accuracy.
248
Results
0.012 T-. . . . . . . . . . . . . . 0010
Minimum sample size for pycnometry
g
The relationship between specific gravity and sample volume as measured by pycnometry is shown by the filled circles in Fig. 2. The dashed line indicates the specific gravity of the homogenate as measured by the density gradient method. Fig. 3 presents the pycnometer measurement variance versus sample volume. Using both of these curves the minimum sample size occurs where the measured specific gravity agrees with the density-gradient value and the measurement variance is at a minimum. Our data indicates that the optimal sample volume is 0.120 ml or larger.
Of I~:: ~! ~
o.oo6
f
i~JLJ ;~C3
o m 0.006 t Z~
>-C) 0 . 0 0 4 t O'-Z
#
0.002
•\ \,
e
\
f
0.000 - ~ - . . . . . . ~ - - - - + - - ~ ............ ~- -F..... 0.000 0.050 0.100 0.150 0.200 0.250 0300 VOLUME [ ml]
Fig. 3. Standard deviation of pycnometry specific-gravity measurement as a function of sample volume.
Accuracy of pycnometer method Fig. 4 displays the accuracy of pycnometry against the calibration standards. The average volume of the standards was 0.120 ml. The slope of the linear regression line was 1.01073, r = 0.99545 which was calculated by least square methods. No significant difference was observed between this slope and 1, the ideal slope ( P < 0.001).
1,300 1 1"260 T
1.18o I
,
Pycnometry L,ersus density gradient column The specific gravity of brain tissue from 5 regions measured with pycnometry and the density-gradient method are shown in Fig. 5; sample size was approximately 0.120 ml and 0.06 ml, respectively. There is no statistically significant
1.0201 1.020
1.050-
1.070
1.040n,(.3 o 1.030.
I 1.14-0
I 1.180
4--1.220
I 1,260
STANDARD SP.GR.
1.060
,ooomo,., Io°61 Den. Gra. Col. n = 6
I 1.100
Fig. 4. Measured specific gravity vs. glass specific-gravity standards,
1.060-
•- -
I 1.060
IS~ Pycnorneter (.n=7) i--7 Den. Gro. Col. ( n = 6 )
1.060 1.050 1.04-0
£3
~m 1.o2o. .
.
.
---~---
.
1.030
1.010.
I .ODD 0.000
1.020 I 0.050
O. 1joO
P 0.150
I 0.200
f 0.250
0.300
VOLUME r rnl ] Fig. 2. S a m p l e v o l u m e vs. specific gravity as m e a s u r e d by p y c n o m e t r y . U n i f o r m density tissue m a t r i x used f o r this study.
1.010 FRO
OCC
MED
PAR
CER
REGION
Fig. 5. Comparison of rat brain specific gravity measured with pycnometry and density-gradient column.
249 100 95
I I~
90 a::
: PYCNOMETRY : DEN.GRA.COL : W / D WEIGHTS
85
245
NSM 111292
Measurement of brain tissue specific gravity using pycnometry G e n e R o b e r t D i R e s t a ~, J o n g b i n L e e 2 and E h u d A r b i t 3 I Nuclear Medicine Research Laboratory, : Neurosurgical Research Laboratoo', and ~ Department Of Surgery, Memorial Sloan-Kettering Cancer Center, New York, NY 10021 (U.S.A.) (Received 10 June 1991) (Accepted 23 July 19911
Key words: Brain density; Water content; Specific gravity; Pycnometry In this paper we introduce and characterize pycnometry, a method used to measure fluid density, fl)r determining a tissue's specific gravity. It uses a 2-ml glass pycnometer filled with distilled water to determine a tissue sample's displacement volume. The tissue's density is determined when it's weight is divided by this volume and specific gravity is computed by dividing the tissue density by the density of water. Pycnometry was validated using pre-calibrated glass, specific gravity standards over the range 1.03-1.26, and compared to the density gradient method using rat brain tissue. We observed that the specific gravity values obtained using pycnometry were highly correlated with the specific gravity standards ( s l o p e - 1.0107, r = I).996) and with the density gradient column when tissue volumes larger than 0.120 ml were used with the pycnometer (slope = 1.0707, r - 0.9826). Good correlation was also observed between percent water content values computed using the Nelson equation with pycnometry or density gradient specific gravity values versus the measured percent water content values obtained with the wet w e i g h t / d r y weight method. Pycnometry is an accurate, reproducible technique to measure tissue specific gravity and brain edema and is best suited for use in a laboratory that engages sporadically in brain edema measurement.
Introduction
Cerebral edema, i.e., increase in brain water content, accompanies a wide variety of pathological processes in the brain and is responsible for the morbidity associated with these conditions. Cerebral edema is particularly prominent in head trauma, stroke, brain tumors, infections and metabolic encephalopathies. The measurement of edema is, therefore, an important indicator used in research in these areas. Edema has traditionally been determined using the wet w e i g h t / d r y weight technique. This method provides a measure of the percent of water content in tissue and is simple to perform
Correspondence: Ehud Arbit, M.D., Memorial SIoan-Kenering Cancer Center, 1275 York Avenue, New York, NY 10021, U.S.A.
(Hossmann et al., 1980; Hiroyuki et al., 1987). Its shortcomings include requiring relatively large samples and long drying times, typically 24 h or more, to ensure reasonable accuracy (Marmarou et al., 1982). Specific gravity measurements provide an alternative index of edema. The technique of using flotation in fluids of known density to determine specific gravity for the determination of serum protein concentrations was first reported by Lowry and Hunter (1945). Laborite and Weber (1965) employed a series of copper sulfate solutions of defined specific gravity to measure changes in edematous rat brains. Nelson et al. (1971) used a gradient column of kerosene and bromobenzene to measure changes in brain water content. The density gradient method was validated by Marmarou et al. (1978) and has since gained wide
246
acceptance among neuroscientists (Fujimoto et al., 1976; Ferszt et al., 1978; Klatzo et al., 1980; Marmarou et al., 1980; Ishige et al., 1987; Karlik and Noseworthy, 1987; Hatashita and Hoff; 1988; Nagao et al., 1988) mainly because of its simplicity and reproducibility, particularly with small samples as compared with the wet weight/dry weight method. The measure of specific gravity lends itself to conversion to percent water content with the Nelson equation (Marmarou et al., 1978; Marmarou et al., 1982). While the density gradient method is an invaluable tool for the study of brain edema, it suffers from several disadvantages. The accuracy of the method depends upon such factors as the linearity of the gradient column, condition of the calibrating standards, temperature and humidity of the dissection area, optimal sample size, and the interval between obtaining the sample and introduction into the density column. The density gradient method uses kerosene and bromobenzene, two volatile, flammable and toxic chemicals, and the column preparation is a time-tedious process. In addition consideration must be given to the tissue's lipid content in relation to its solubility in the gradient column's solutions. Lastly, errors are introduced into the measuremcnt of brain tissue water when the edema fluid contains protein (Steward-Wallace, 1939; Shigeno et al., 1982; Bothe and Hossman, 1984). In this paper we introduce pycnometry, a simple method frequently used to determine fluid density, to measure tissue specific gravity (Kitagawa et al., 1981). We evaluate its accuracy against glass specific gravity standards, determine the optimal specimen size using a uniform density material, and correlate it against both the column density gradient and wet weight/dry weight methods using rat brain tissue specimens.
Materials and methods
Pycnometer method The pycnometer method requires a micro-analytical balance, i.e., 0.01 mg resolution, and a small 2-ml glass pycnometer (Fisher Inc.) shown in Fig. 1. Its center hole was plugged with water-
PYCNOMETER
-~'~/~mar~ks ~...
Alignment ~ - ~ ' ~
Epoxy ~ugs
FLASK (2 ml) Fig. I. Standard 2-ml glass pycnometer (Fisher Scientific). Black marks placed on exterior of flask and on ground glass stopper; waterproof epoxy used to seal both stopper ends.
proof epoxy to minimize evaporative losses. Two black marks were placed on the ground-glass stopper and on the outside neck of the flask to ensure equivolume between measurement. The pycnometer was carefully cleaned prior to use and wiped dry using lint-free paper. The fluid used in our studies was distilled water which had been filtered and degassed by placing it under vacuum for 1 h prior to use. The method is used as follows with all measurements made at thc same temperature. (a) The pycnometer is filled with pure distilled water which has reached the desired temperature, i.e., 23 ° C. The stopper is introduced and its mark is aligned with the pycnometer mark. The outside of the flask is thoroughly dried with lintfree wipes and weighed (mot.). Air bubbles, if present, are eliminated by removing the stopper and immersing the flask into an ultrasonic cleaner for several seconds. (b) The tissue sample(s) is weighed. (c) Some water from the pycnometer is removed and the tissue transferred into the flask. The pycnometer is refilled with water, bubbles are removed, marks aligned and the outside of the flask is dried thoroughly. The pycnometer containing the sample is then weighed ( m r ) . (d) The tissue specific gravity is calculated by dividing the tissue density (derived from Eqn. t below) by the density of water at the measurement temperature (d,,).
247
Eqn. 1 was used to determine tissue density from the weight measurements (its derivation is presented in the Appendix):
d~=
S
(m~j-m~ +s)
xdw
were re-weighed. The difference between the tare weight and the new weight was the tissue's 'dry weight'. The percent water content is determined from Eqn. 2 below:
(1)
wet weight-dry weight % water content =
where d~ = density of the sample (g/ml), d,, = density of water ( g / m l ) at temperatures of measurement, s = weight of sample (g), mpj,= weight of pycnomcter and water (g), m t = weight of pycnometcr, water and sample (g). Now specific gravity is related to density by the relationship: specific gravity = d J d w.
Density gradient column method The density gradient single column system, Model DC-1, Techne Inc., was used in this study. The density gradient was prepared with kerosene and bromobenzene according to the method presented by Marmarou et al. (1982). The column of liquid was allowed to reach a temperature of 2 3 ° C and then calibrated using glass specific gravity standards (Techne Corporation) over the range 1.03-1.26. With a properly prepared gradient column, the specific gravity varies linearly according to the height of the column. The standards' positions within the column versus their specific gravity was plotted and evaluated for its linearity using statistical regression analysis. The column was accepted if the correlation coefficient was at least 0.9995; if not, the fluids were discarded and the gradient re-established. Specimens weighing 60 mg were lowered into the column at 5-min intervals and allowed to descend. The specific gravity of the sample was determincd from the depth of the sample read at 5 rain post-immersion. This position was related to the sample's specific gravity with the calibration plot. Wet weight/dry weight method Aluminum specimen dishes were labeled and tared prior to dissection of brain and again immediately after receiving its sample. The difference between these 2 weights is the tissue's 'wet weight'. The dishes were then placed into a drying oven adjusted to 9 0 ° C . After 5 days, they
wet weight
x 100
(2)
Specific grat,ity to percent water content com,ersion method Specific gravity values measured by either the density gradient or pycnometry were converted into percent water content using Nelson's method (Nelson et al., 1971) This technique relates spccific gravity to its approximate (g:g) H 2 0 / t i s s u e value. Surgical procedures and tissue preparations S p r a g u e - D a w l e y male rates (300-400 g) were anesthetized with a mixture of 5 m g / k g xylazine and 80 m g / k g ketamine and then killed by an overdose injection of saturated KCI. The rat brain was then quickly removed from the cranium. Large tissue samples were taken from the frontal, parietal, occipital cortex, cerebellum and medulla. Each sample was divided into two pieces, one weighing approximately 120 mg for the pycnometer or the w e t / d r y method and the second weighing approximately 60 mg for the density gradient method. Samples were stored in pre-chillcd petri dishes until used. Uniform density tissue preparation The homogenized tissue from a whole rat brain was mixed with 10 ml of warm, sterile 30% gelatine solution. The mixture was poured into a sterile petri dish, covered, refrigerated and allowed to set. Its density was determined using a 60 mg sample with the column density gradient method. The remaining matrix was cut into samples of varying sizes. The density of these samples was determined using the pycnometer method to determine the influence of sample volume on pycnometry accuracy.
248
Results
0.012 T-. . . . . . . . . . . . . . 0010
Minimum sample size for pycnometry
g
The relationship between specific gravity and sample volume as measured by pycnometry is shown by the filled circles in Fig. 2. The dashed line indicates the specific gravity of the homogenate as measured by the density gradient method. Fig. 3 presents the pycnometer measurement variance versus sample volume. Using both of these curves the minimum sample size occurs where the measured specific gravity agrees with the density-gradient value and the measurement variance is at a minimum. Our data indicates that the optimal sample volume is 0.120 ml or larger.
Of I~:: ~! ~
o.oo6
f
i~JLJ ;~C3
o m 0.006 t Z~
>-C) 0 . 0 0 4 t O'-Z
#
0.002
•\ \,
e
\
f
0.000 - ~ - . . . . . . ~ - - - - + - - ~ ............ ~- -F..... 0.000 0.050 0.100 0.150 0.200 0.250 0300 VOLUME [ ml]
Fig. 3. Standard deviation of pycnometry specific-gravity measurement as a function of sample volume.
Accuracy of pycnometer method Fig. 4 displays the accuracy of pycnometry against the calibration standards. The average volume of the standards was 0.120 ml. The slope of the linear regression line was 1.01073, r = 0.99545 which was calculated by least square methods. No significant difference was observed between this slope and 1, the ideal slope ( P < 0.001).
1,300 1 1"260 T
1.18o I
,
Pycnometry L,ersus density gradient column The specific gravity of brain tissue from 5 regions measured with pycnometry and the density-gradient method are shown in Fig. 5; sample size was approximately 0.120 ml and 0.06 ml, respectively. There is no statistically significant
1.0201 1.020
1.050-
1.070
1.040n,(.3 o 1.030.
I 1.14-0
I 1.180
4--1.220
I 1,260
STANDARD SP.GR.
1.060
,ooomo,., Io°61 Den. Gra. Col. n = 6
I 1.100
Fig. 4. Measured specific gravity vs. glass specific-gravity standards,
1.060-
•- -
I 1.060
IS~ Pycnorneter (.n=7) i--7 Den. Gro. Col. ( n = 6 )
1.060 1.050 1.04-0
£3
~m 1.o2o. .
.
.
---~---
.
1.030
1.010.
I .ODD 0.000
1.020 I 0.050
O. 1joO
P 0.150
I 0.200
f 0.250
0.300
VOLUME r rnl ] Fig. 2. S a m p l e v o l u m e vs. specific gravity as m e a s u r e d by p y c n o m e t r y . U n i f o r m density tissue m a t r i x used f o r this study.
1.010 FRO
OCC
MED
PAR
CER
REGION
Fig. 5. Comparison of rat brain specific gravity measured with pycnometry and density-gradient column.
249 100 95
I I~
90 a::
: PYCNOMETRY : DEN.GRA.COL : W / D WEIGHTS
85
